Adonis Diaries

Archive for the ‘professional articles’ Category

Inspiring Young Inventors? Not investors, please…

An “experimental learning workshop” where kids engage in an essential but increasingly rare activity: they make stuff.

 posted this November 25, 2013 on Mind/Shift

How Do We Inspire Young Inventors?

In New Haven, Connecticut, where I live with my husband and two sons, we are lucky to have nearby the Eli Whitney Museum.

This place is the opposite of a please don’t touch repository of fine art. It’s an “experimental learning workshop” where kids engage in an essential but increasingly rare activity: they make stuff.

Looking around my living room, I can see lots of the stuff made there by my older son: a model ship that can move around in water (in solid ice is more relevant for those trapped in the Arctic) with the aid of a battery-powered motor he put together; a “camera obscura” that can project a real-world scene onto a wall in a darkened room; a wooden pinball game he designed himself. (You can view an archive of Eli Whitney Museum projects here.)

The people who run Eli Whitney call these hands-on projects “experiments.” As they put it:

“Experiments are a way of learning things. They require self-guided trial and error, active exploration, and testing by all the senses.

Experiments begin with important questions, questions that make you think or that inspire you to create.”

This process of exploring, testing and finding out is vital to children’s intellectual and psychological development—but opportunities to engage in it are fewer than they once were.

Frank Keil, a Yale University psychologist who is in his early 60′s said: “My friends and I grew up playing around in the garage, fixing our cars. Today kids are sealed in a silicon bubble. They don’t know how anything works.”

“We scour the country looking for young builders and inventors. They’re getting harder and harder to find.”

Many others have noticed this phenomenon.

Engineering professors report that students now enter college without the kind of hands-on expertise they once unfailingly possessed.

Kim Vandiver, dean for undergraduate research at the Massachusetts Institute of Technology said:

“We scour the country looking for young builders and inventors. They’re getting harder and harder to find.” MIT now offers classes and extracurricular activities devoted to taking things apart and putting them together, an effort to teach students the skills their fathers and grandfathers learned curbside on weekend afternoons.

Why should this matter?

Some would argue that the digital age has rendered such technical know-how obsolete.

Our omnipresent devices work the way we want them to (well, most of the time), with no skill required beyond pushing a button. What’s to be gained by knowing how they work?

Actually, a lot.

Research in the science of learning shows that hands-on building projects help young people conceptualize ideas and understand issues in greater depth.

In an experiment described in the International Journal of Engineering Education in 2009, for example, one group of eighth-graders was taught about water resources in the traditional way: classroom lectures, handouts and worksheets.

Meanwhile, a group of their classmates explored the same subject by designing and constructing a water purification device. The students in the second group learned the material better: they knew more about the importance of clean drinking water and how it is produced, and they engaged in deeper and more complex thinking in response to open-ended questions on water resources and water quality.

If we want more young people to choose a profession in one of the group of crucial fields known as STEM—science, technology, engineering and math—we ought to start cultivating these interests and skills early.

But the way to do so may not be the kind of highly structured and directed instruction that we usually associate with these subjects. Instead, some educators have begun taking seriously an activity often dismissed as a waste of time: tinkering.

Tinkering is the polar opposite of the test-driven, results-oriented approach of No Child Left Behind: it involves a loose process of trying things out, seeing what happens, reflecting and evaluating, and trying again.

“Tinkering is the way that real science happens, in all its messy glory,” says Sylvia Martinez, co-author of the new book Invent To Learn: Making, Tinkering, and Engineering in the Classroom.

Martinez is one of the leaders of the “makers’ movement,” a nationwide effort to help kids discover the value of getting their hands dirty and their minds engaged.

The next generation of scientists—and artists, and inventors, and entrepreneurs—may depend on it.

Note: Read my articles in category Human Factors Engineering on Teaching methods https://adonis49.wordpress.com/2008/10/26/teaching-methods/

Advertisements

 

45 top examples of letterpress business cards

Using the oldest form of printing to create memorable letterpress business cards can help beat off the competition. Here are some examples of exactly what can be achieved.

 hot metal letters
Letterpress business cards will surely get you noticed

Letterpress printing has been around for hundreds of years and yet remains one of the hottest trends in stationery. Letterpress business cards still have the ability to turn heads.

The ancient technique involves a surface with raised letters or artwork inked and then pressed into thick, soft paper, adding an exciting, tactile quality to the design.

Known for producing a clean and elegant style, the letterpress technique is favoured by many designers for creating unique and creative business cards.

Here we’ve picked 45 brilliant examples of letterpress cards: which one is your favourite?

01. Oddds

Beautiful letterpress buisness cards for a Singapore based agency

Combining craft and precision to produce a beautiful aesthetic, these new cards from Singapore based agency Oddds are about as beautiful as they come.

“The identity broadens with usage of letterpress accompanied by glamour with modernism,” they explain. “This is created with the intertwining between bronze and a soft tone of turquoise. The play of metal tones and pieces with cotton represents the designer’s ideologies of art direction and design.”

02. Elias Mendoza

These classic-looking cards aim to convey the unique and personalised touch that Mendoza brings to his work

When Cocoa Branding was hired to create branding materials for Elias Mendoza, one of the most prominent immigration and naturalisation attorneys in the United States, it needed to convey the unique and personalised touch that Mendoza brings to his work with all types of family and business immigration needs. Part of the branding solution, designed by Rodrigo Suárez, are these heavyweight letterpress business cards.

03. Featherweight Printing

These weighty cards feature bold edge painting to really help them stand out

Creating business cards for her printing company has been a learning experience for Kate Huat. After a first run that didn’t turn out quite as hoped, Kate invested in some Crane Letra 200lb stock that is hefty enough to take a front imprint that won’t ruin the pattern on the back of the car, and that’s also suitable for bold edge painting that really stands out. “I was a little nervous to spray these,” says Kate, “since I had spent so much time and care on each step along the way, but I am very pleased with the outcome and hope to have more edge painting opportunities soon.”

04. Bailey H. Robinson

Two Arms Inc aimed for a turn-of-the-century look with this card, and got it spot on

Brooklyn-based tattoo artist Bailey H. Robinson has a very distinct take on traditional American tattooing, and when he brought in Two Arms Inc to produce a set of custom letterpress business cards, they were keen to design something that looked and felt like traditional turn-of-the-century type. The resulting card is an ornate delight that wouldn’t have looked out of place in 1901 (except for the contact URL, of course).

05. Two Sisters Photography

These lush cards take their inspiration from Venn diagrams and cameos

Commissioned to create an identity for Two Sisters PhotographyFizz took inspiration from a Venn diagram and used that as the basis for cameo-style silhouettes. The silhouettes enabled Fizz to showcase each sister’s style and personality through their profile treatment, and alongside the Venn-style card Fizz also created individual designs for each sister. Fizz went all-out in the production stakes, using three Pantone colours on 220lb Crane Letra cotton stock, and finishing everything off with a custom die-cut.

Each sister has her own individual design to go with the striking main design 

06. Impressworks

Typographer Mia Parcell gets a circular letterpress card

Impressworks printed these business cards for designer Mia Parcell. The job features a two-colour, double-sided print on our 600gsm Fluoro White Crane Lettra stock. Mia took care of the artwork herself, with Impressworks coming up with the circular execution.

07. Simon Featherstone

A contemporary logo design and letterpress card for the lighting designer

Simon Featherstone is a freelance lighting designer & programmer with vast experience in video and LED technology. He needed a branding refresh to help portray the qualities and capabilities of his practice to a wider audience.

The contemporary logo marque used a pattern derived from LEDs and combined it with a bold gradient running from cyan to magenta. The branding collateral uses a range of cool grey tones to suitably reflect the high-end nature of the industry and communicate a professional and knowledgable tone of voice.

08. Marilyne Scheepers

Simple typography helps make this card design a winner

Marilyne Scheepers created this letterpress business card designGraphic Wand studio. The turquoise colouring on the edge of the cards enables a striking effect when staked. Simple typography and a clever logo allow these letterpress business cards to speak for themselves.

09. Adicto

Combining authentic design with passionate craftsmenship

When it comes to paper, Gmund relies on tradition and innovation. On the oldest still producing paper machine in Europe, standards are set for the finest papers. Offizin Parnassia is a fine art studio born out of a love for old books. These Adictocards were created to combine authentic design with passionate craftsmenship.

10. Lucky Cat Acupuncture

A adorable anime influenced letterpress business card

Identity for designer Lovely Mpls‘ wife’s acupuncture clinic in Minneapolis. “She requested something that had a little bit of an anime style with some whimsy,” he explains. The business cards were letterpress printed at Studio on Fire.

11. Pepelatz

 Andrej Barmelaj Designer and illustrator Andrej Barmelaj developed these postage-stamp inspired letterpress business cards for advertising agency Pepelatz

Ukraine-based designer and illustrator Andrej Barmelaj is the man behind these brilliant postage stamp-inspired letterpress business cards for independent advertising agency Pepelatz.

With four different designs to choose from, each stamp adds a splash of colour to Pepelatz employees otherwise simple white letterpress business card.

Letterpress business cards
The postage stamp-inspired designs add a touch of colour to these creative business cards

12. German Torres

 German TorresLetterpress business cards for illustrator German Torres feature his character transforming from human to werewolf

These beautiful letterpress business cards feature the wild illustrator German Torres transforming from human to werewolf. Print shop La Trasteria created them using a two colour split fountain for the transformation and black ink for the other details.

These creative business cards are beautifully printed, and we just love the playful design, which depicts a smiley character by day and the stages that turn him into a wild illustrator by night.

 La Trasteria
Print shop La Trasteria created gorgeous letterpress business cards from these brilliant illustrations

13. Dane Holmquist

13. Dane Holmquist

 Dane Holmquist
American artist Dane Holmquist created the illustration for these letterpress business cards himself

These detailed, beautiful letterpress business cards belong to illustrator and graphic designer Dane Holmquist. The talented artist created the intricate illustration himself, before elisting the help of printers DSJ in Santa Monica, who helped bring his vision to life.

Holmquist’s mantra is ‘If it has the potential to be cool, I will do my very best to make it so’. And he’s certainly achieved that with this gorgeous design.

14. Jee

 Johanna Ecker'
Johanna Ecker’s gorgeous letterpress business cards double as a tiny notebook

Brandconsultant Johanna Elisabeth Ecker wanted a business card that reflected both her personality and quality of her work.

So, working alongside designer Kurt Glanzer at Moodley Brand Identity, she developed these beautiful letterpress business cards, which doubles as a tiny notebook, which include a personal handwritten message for the receiver and 15 more empty pages for them to use as they wish.

Letterpress business cards
Each letterpress business card includes a personal handwritten message from Ecker

15. Bentply

Letterpress business cards
One minute it’s a business card, the next it’s a Gerald Summers armchair. Cool, huh?

What you see here is one of the neat letterpress business cards for furniture shop Bentply in Marylebone, London that can be made into mini bent plywood chairs.

The ingenious card design was created by art director, writer, and designer Richard C Evans and produced by Elegante Pressin Lithuania. Just follow the instructions and the kiss-cut card can be folded into a miniature of the iconic 1934 plywood armchair designed by Gerald Summers.

Creative business cards
All Elegante’s products are hand made using 100% cotton (tree-free) paper and oil based inks

16. Dot Design

letterpress business cards
The distinct shape makes these creative business cards instantly striking

Printed using two colours on both sides and die cut to shape, the carefully placed design elements on each side of these letterpress business cards avoid any overlapping. They’re printed on 425gsm cotton and are instantly striking. Creating such a distinct shape really makes these creative business cards one of our favourites.

17. Derek Welsh

letterpress business cards
The sleek finish represents Derek’s work perfectly

Glasgow Press achieved a monster 810gsm when creating these letterpress business cards by bonding 270gsm sheets of Ebony, Citrine and Bright White to achieve a stand out card for furniture maker Derek Welsh. Black letterpress to the Bright White side with the gloss black foil to the Ebony card really represent, making this a unique creation.

letterpress business cards
The gloss foil finish really makes these creative business cards stand out

18. Whitney Shaw

letterpress business cards
We love the colours used on this letterpress business card

The colours really won us over with this letterpress business card. Encorporating a variety of fonts, as well as a playful illustration, this is a design that would surely be cherished by the receiver. The use of larger ‘W’s’ also allow for Whitney’s name to stick in the mind, whilst keeping the design aspects to a minimal affect.

19. Bryon Darby

letterpress business cards
The clean-cut, minimal approach works for this letterpress business card

Often, less is more when it comes to design. It’s certainly the case for this letterpress business card for photographer Bryon Darby. Using a minimal colour scheme and using the letterpress to create a series of beautiful patterns, we love its simplicity.

20. For Luca

letterpress business cards
A lovely letterpress card for a worthy cause

When Luca was just three years old he was struck down with Meningococcal Septicaemia – losing both of his legs along with muscle and skin below his hips. When his dad contacted Blush Publishing for some business cards to help with Luca’s campaign they fired up the presses and produce this delightful design. We love the bright blue typography placed on a simple, white backdrop. Instantly eye-catching.

21. Amy Weibel

The typography is simple but effective

Amy Weibel is a art director & designer, with a passion for food, travel, technology and of course design. She is currently based in the bustling neighborhood of Chinatown, New York. The typography treatment on each side is simple and effective; whilst the trim is bold and memorable.

The trim is bold and memorable

22. Forge

The wraparound sticker is an inspired touch

A self-promotional business card for the design studio Forge. “We wanted a card that would communicate our hands-on and unique approach to design solutions,” they explain on their site. “We sourced the paper ourselves and used two printers.” The offset typography is pleasing, printed on a 200lb uncoated stock, with the wraparound sticker an inspired touch.

23. Kitty’s Beauty Parlour

The card was letterpress-printed on double thick 236lb cotton card stock

Print & Grain is full service graphic design and letterpress studio. They offer original custom design and letterpress printing. Print & Grain encourage a collaborative effort with their clients during the design process to create a unique and personal card, which will be letterpress printed on double thick 236lb cotton card stock, using an antique Kelsey Excelsior printing press. A strong example from their portfolio is this creation for Kitty’s Beauty Parlour.

The studio uses an antique Kelsey Excelsior printing press

24. Mei Yen Chua

The card combines slab-serif and serif typefaces

A variety of techniques were used on this business card for Mei Yen Chua. The experimental techniques using overlays of ink works wonders. Each shape also been embossed into the card, combining slab-serif and serif typefaces. A confident and eye-catching card.

25. Moglea

We love the incorporation of natural images into this card

Moglea is a boutique letterpress stationery design studio created by Meg Gleason. She loves bold pops of colour, hand-lettered typography and intricate floral and geometric patterns. The letterpress studio is within a farmhouse in Western Lowa, and this influences her style, which incorporates animals and nature into the designs.

26. Denim Geek

letterpress business cards
Blush pull off yet another beautiful letterpress printing job

This gorgeous letterpress business card comes from those talented folks at Blush – a bespoke and custom letterpress printing service that offers 
wedding stationery, personalised correspondence, greeting cards and of course, business cards.

Even though white letterpress ink is often extremely diffcult to work with and doesn’t give a very good solid on dark stock, it works extremely well in this design. The card was printed on 1000mic recycled board, so it’s good for the planet too!

27. Leigh Cameron

letterpress business cards
We love the stark contrast within the colour choices

The guys over at Typoretum have an incredibly impressive array of letterpress business card designs on offer. Specialising in comprehensive letterpress business card and stationery design, the printing service can work from artwork supplied or help you to design your own.

This creation for Leigh Cameron is simple, elegant and engaging. We love the stark contrast of the green coupled with the grey. The cards are printed on 750 micron recycled greyboard.

28. Britt Boyd

letterpress business cards
Combining stunning graphic design and typography, this is one of their best examples

Print and Grain are quickly becoming known for their letterpress business card offerings within the design industry. Based in Portland, Oregon, they also provide letterpress inspired greetings cards through their Etsy shop.

This design for Britt Boyd is a stunning feat in typography and graphic design. The beautiful graphics, combined with an eye-catching font and coloured edges, make it one of their best examples.

29. Ninja star

letterpress business cards
A letterpress business card that will get ‘the point’ across

Now, here’s a letterpress business card that’ll catch a potential client’s eye for all the right reasons. Showcased on Beast Pieces, this design will certainly get ‘the point’ across (sorry!) Although it might not be the ideal shape and size to slip into your wallet, the understated letterpress design and carefully placed typography make it a must in this list.

30. Nili Studios

letterpress business cards
Print & Grain triumph again with this letterpress design for Nili Studios

Another addition from the guys at Print & Grain, this lovely-looking letterpress business card was designed for creative agency Nili Studios. The nautical-inspired theme is perfectly executed with the chosen soft blues as well as the waves and anchor logo. The choice of font is also a brilliant choice, as its readability is showcased.

31. Druckerei Eisenhardt

letterpress business cards
The bold choice in size, shape and typography make this business card stand out

A printing press that’s been operating for over 42 years,Eisenhardt is a trusted German company offering the best in letterpress design. A family-run business, it works with both modern and traditional techniques in the centre of Frankfurt.

This letterpress business card design immediately caught our eye thanks to its bold choice in size, shape and typography. The orange-on-white colour scheme enables the font to stand out whilst the understated letterpress gives it the edge it needs.

32. Stu Horvath

letterpress business cards
Dolce Press showcase the intensity of one-colour printing

Dolce Press is a boutique print shop and design studio located on the eastern end of Long Island. Specialising in letterpress, the design studio creates one-of-a-kind prints.

This letterpress business card was printed on super-thick kraft stock in black ink, showcasing what can be accomplished with one-colour printing. The letterpress cards were finished off with black edge colouring and printed on a Chandler & Price 8×12.

33. Silk Lab

letterpress business cards
We love Lettera Magica’s spin on the usual rectangular shape

A workshop based in Cracow, Lettera Magica combined their love for tradition, perfection, patience and respect for nature to create some of the most stunning letterpress business cards we’ve ever seen.

This design was crafted for creative studio Skill Lab and was printed on Duplex Orange paper. It measures at 90 x 55mm, which makes it the perfect size to slip into your wallet. We love their slight change in the usual rectangular offering.

34. J. Fletcher Design

letterpress business cards
J. Fletcher perfectly echoes his client’s ethos, with this natural-looking offering

J. Fletcher has been producing graphic solutions for clients ever since 2004. Describing himself as a ‘one-man show, not a one-man band’ his executions in branding are faultless. This design – created for Charleston Naturally – perfectly captures their ethos.

They’re made from the highest quality products – often organic, mindfully local. Fletcher echoes this ‘natural’ ethos with this beautiful letterpress business card design.

35. Jason Turiff

letterpress business cards
The ink used for Jason’s business card was the perfect balance of blue and green

Another addition from Dolce Press; this time, for studio engineer Jason Turiff. Every colour here is hand mixed from a base of 16 colors which allows Dolce Press to achieve the full spectrum of Pantone colours and more. The ink used for Jason’s business cards was the perfect balance of blue and green. They were then printed on Crane’s Lettra (Pearl White, 300gsm) paper.

36. Fizz

Letterpress printing
These coasters complement Fizz Creative’s ‘Design straight up with a twist’ motto perfectly

Graphic designers Jasen Melnick and Katie Major used letterpress printing to create these awesome retro-style coasters to promote their award-winning boutique studio Fizz Creative. Their letterpress business cards design perfectly complements the duo’s motto, ‘Design straight up with a twist’.

Featuring two Pantone colours, and printed on thick, ivory white stock, these business cards won awards from the AIGA Cleveland Design Competition and from the HOW+Print Color in Design Competition.

37. Dare

Letterpress printing
These letterpress printed cards feature ASCII images of individual employees. We want some!

We love these innovative designs by creative agency Dare. The company describes itself as ‘a merger of a traditional agency and digital shop’, so it wanted letterpress business cards that also combine old with new. Employees were invited to upload a picture of themselves and enter their email and phone numbers. An ASCII image was then automatically created using characters from their contact details.

The company then chose to letterpress these images onto individual business cards in order to re-emphasise the idea of traditional techniques meeting digital. Awesome.

38. Pablo Abad

Letterpress printing
Pablo Abad created this sans-serif typeface specifically for this project

This beautiful design was created by Madrid-based graphic designer and art director Pablo Abad. In a quest to create a new visual identity, Abad designed a bespoke geometric sans-serif typeface gara.

The clean, yet edgy letterpress business cards design makes it eye-catching but still easy to understand. Abad’s love of illustration and typography is clear in this very personal project.

39. Shyama Golden

Letterpress printing
Shyama Golden – ‘It Rhymes with Llama’. As in Shama-Llama, get it?

It’s always an awkward moment when you can’t pronounce someone’s name or get it wrong. To ensure that never happens again, artist and designer Shyama Golden designed these funky letterpressed business cards with her catchphrase ‘It Rhymes with Llama’.

Printed on Cranes lettra cotton paper, with 100% hand-drawn type and a hand-drawn llama, these beauties are one-colour letterpress printing at its best.

40. Luke Lucas

Letterpress printing
Luke Lucas’s card uses specialist ink which makes it glow in the dark. Ooooooooh!

After developing a new script-based logo, typographer and graphic designer Luke Lucas decided to update his business cards to show it off. Featuring his name on one side and his logo on the other, Lucas’s card is no ordinary one.

Printed on 600gsm Lettra cotton stock, this letterpress business cards design uses phosphorescent ink so that in regular light it appears as a simple debossing but when the lights dim, the logo illuminates in a bright green. Very cool.

41. James Prunean

Letterpress printing
James Prunean lovingly created these awesome business cards for his brother Ovi

As is often the case, simplicity can work wonders. A perfect example of this is these gorgeous letterpressed business cards created by graphic designer James Prunean for his brother, Ovi.

Prunean used strong, bold colours that immediately catch your eye and at a first glance look very similar to Pantone colour cards. Printed on 220lb cotton paper, the soft yet strong impression has reportedly helped painter Ovi beat off a lot competition. And we can see why!

42. Jennifer Daniel

Letterpress printing
Jennifer Daniel loves unicorns. Can you tell?

These quirky but cool business cards belong to Bloomberg Buisnessweek’s graphic director Jennifer Daniel. An internationally recognised illustrator, designer and art director, Daniel is also a lover of unicorns.

After having a hard time choosing from her designs, she decided to have three printed, each one as memorable as the next and featuring her RSI-inducing contact email along the bottom. We wonder whether she’s got any unicorn grooming business off the back of them?

43. Plaid Lab

Letterpress printing
Does anyone else want to stroke this business card?

The Plaid Lab is the portfolio of designer Mark Saunders, and the design for his business card was clearly inspired by that name. The gorgeous, brightly coloured letterpress business cards design features a plaid pattern, consisting of crossed horizontal and vertical bands in two or more colours.

The cards were letterpress printed on heavy 220lb Cover Crane Lettra Pearl White stock and used overprinting inks to create the rich and saturated colours. The printing is so good, it actually looks like Plaid material forms part of the card. An obvious but very clever design.

44. Rabbit Hole

Letterpress printing
With this business card you also get a cute desktop companion

This beautiful design grabbed our attention as soon as we saw it. Why? Not only because of the gorgeous impressions in fluourescent green with a double hit of silver on Beer Matt Board 390gsm but also because you can turn it into a rabbit.

Yes, you read that right. Several shapes are die-cut out and can be put together to create a stylised bunny. Cool, huh? We bet the guys at letterpress and design agency The Hungry Workshop had great fun creating these for The Rabbit Hole Ideation Cafe.

45. Still Liquor

Letterpress printing
The message here is not to drink and drive. Hence the word still.

This classy design was created for Still Liquor, a bar and micro distillery in a reclaimed bootleg liquor and auto shop in Seattle. To convey the bar’s rich heritage, New York-based multi-disciplinary design practice Javas Lehn studio decided to use the iconic Ford Model T silhouette, bold type and unexpected contemporary compositions.

The final, clean design was then letterpress-printed onto Eska board stock, using a bright shade of red to accentuate the bold type. If the bar itself is a classy as its business card, we’d definitely drink there.

Have you seen a great example of letterpress business cards? Let us know in the comments!

 hot metal letters
Letterpress business cards will surely get you noticed

Letterpress printing has been around for hundreds of years and yet remains one of the hottest trends in stationery. Letterpress business cards still have the ability to turn heads.

The ancient technique involves a surface with raised letters or artwork inked and then pressed into thick, soft paper, adding an exciting, tactile quality to the design.

Known for producing a clean and elegant style, the letterpress technique is favoured by many designers for creating unique and creative business cards. Here we’ve picked 44 brilliant examples of letterpress cards: which one is your favourite?

01. Oddds

letterpress business cards
Beautiful letterpress buisness cards for a Singapore based agency

Combining craft and precision to produce a beautiful aesthetic, these new cards from Singapore based agency Oddds are about as beautiful as they come.

“The identity broadens with usage of letterpress accompanied by glamour with modernism,” they explain. “This is created with the intertwining between bronze and a soft tone of turquoise. The play of metal tones and pieces with cotton represents the designer’s ideologies of art direction and design.”

02. Elias Mendoza

letterpress business cards
These classic-looking cards aim to convey the unique and personalised touch that Mendoza brings to his work

When Cocoa Branding was hired to create branding materials for Elias Mendoza, one of the most prominent immigration and naturalisation attorneys in the United States, it needed to convey the unique and personalised touch that Mendoza brings to his work with all types of family and business immigration needs. Part of the branding solution, designed by Rodrigo Suárez, are these heavyweight letterpress business cards.

03. Featherweight Printing

letterpress business cards
These weighty cards feature bold edge painting to really help them stand out

Creating business cards for her printing company has been a learning experience for Kate Huat. After a first run that didn’t turn out quite as hoped, Kate invested in some Crane Letra 200lb stock that is hefty enough to take a front imprint that won’t ruin the pattern on the back of the car, and that’s also suitable for bold edge painting that really stands out. “I was a little nervous to spray these,” says Kate, “since I had spent so much time and care on each step along the way, but I am very pleased with the outcome and hope to have more edge painting opportunities soon.”

04. Bailey H. Robinson

letterpress business cards
Two Arms Inc aimed for a turn-of-the-century look with this card, and got it spot on

Brooklyn-based tattoo artist Bailey H. Robinson has a very distinct take on traditional American tattooing, and when he brought in Two Arms Inc to produce a set of custom letterpress business cards, they were keen to design something that looked and felt like traditional turn-of-the-century type. The resulting card is an ornate delight that wouldn’t have looked out of place in 1901 (except for the contact URL, of course).

05. Two Sisters Photography

letterpress business cards
These lush cards take their inspiration from Venn diagrams and cameos

Commissioned to create an identity for Two Sisters PhotographyFizz took inspiration from a Venn diagram and used that as the basis for cameo-style silhouettes. The silhouettes enabled Fizz to showcase each sister’s style and personality through their profile treatment, and alongside the Venn-style card Fizz also created individual designs for each sister. Fizz went all-out in the production stakes, using three Pantone colours on 220lb Crane Letra cotton stock, and finishing everything off with a custom die-cut.

letterpress business cards
Each sister has her own individual design to go with the striking main design

06. Impressworks

letterpress business cards
Typographer Mia Parcell gets a circular letterpress card

Impressworks printed these business cards for designer Mia Parcell. The job features a two-colour, double-sided print on our 600gsm Fluoro White Crane Lettra stock. Mia took care of the artwork herself, with Impressworks coming up with the circular execution.

07. Simon Featherstone

examples of letterpress business cards
A contemporary logo design and letterpress card for the lighting designer

Simon Featherstone is a freelance lighting designer & programmer with vast experience in video and LED technology. He needed a branding refresh to help portray the qualities and capabilities of his practice to a wider audience.

The contemporary logo marque used a pattern derived from LEDs and combined it with a bold gradient running from cyan to magenta. The branding collateral uses a range of cool grey tones to suitably reflect the high-end nature of the industry and communicate a professional and knowledgable tone of voice.

08. Marilyne Scheepers

examples of letterpress business cards
Simple typography helps make this card design a winner

Marilyne Scheepers created this letterpress business card designGraphic Wand studio. The turquoise colouring on the edge of the cards enables a striking effect when staked. Simple typography and a clever logo allow these letterpress business cards to speak for themselves.

09. Adicto

letterpress business cards
Combining authentic design with passionate craftsmenship

When it comes to paper, Gmund relies on tradition and innovation. On the oldest still producing paper machine in Europe, standards are set for the finest papers. Offizin Parnassia is a fine art studio born out of a love for old books. These Adictocards were created to combine authentic design with passionate craftsmenship.

10. Lucky Cat Acupuncture

examples of letterpress business cards
A adorable anime influenced letterpress business card

Identity for designer Lovely Mpls‘ wife’s acupuncture clinic in Minneapolis. “She requested something that had a little bit of an anime style with some whimsy,” he explains. The business cards were letterpress printed at Studio on Fire.

11. Pepelatz

 Andrej Barmelaj
Designer and illustrator Andrej Barmelaj developed these postage-stamp inspired letterpress business cards for advertising agency Pepelatz

Ukraine-based designer and illustrator Andrej Barmelaj is the man behind these brilliant postage stamp-inspired letterpress business cards for independent advertising agency Pepelatz.

With four different designs to choose from, each stamp adds a splash of colour to Pepelatz employees otherwise simple white letterpress business card.

Letterpress business cards
The postage stamp-inspired designs add a touch of colour to these creative business cards

12. German Torres

 German Torres
Letterpress business cards for illustrator German Torres feature his character transforming from human to werewolf

These beautiful letterpress business cards feature the wild illustrator German Torres transforming from human to werewolf. Print shop La Trasteria created them using a two colour split fountain for the transformation and black ink for the other details.

These creative business cards are beautifully printed, and we just love the playful design, which depicts a smiley character by day and the stages that turn him into a wild illustrator by night.

 La Trasteria
Print shop La Trasteria created gorgeous letterpress business cards from these brilliant illustrations

13. Dane Holmquist

 Dane Holmquist
American artist Dane Holmquist created the illustration for these letterpress business cards himself

These detailed, beautiful letterpress business cards belong to illustrator and graphic designer Dane Holmquist. The talented artist created the intricate illustration himself, before elisting the help of printers DSJ in Santa Monica, who helped bring his vision to life.

Holmquist’s mantra is ‘If it has the potential to be cool, I will do my very best to make it so’. And he’s certainly achieved that with this gorgeous design.

14. Jee

 Johanna Ecker'
Johanna Ecker’s gorgeous letterpress business cards double as a tiny notebook

Brandconsultant Johanna Elisabeth Ecker wanted a business card that reflected both her personality and quality of her work.

So, working alongside designer Kurt Glanzer at Moodley Brand Identity, she developed these beautiful letterpress business cards, which doubles as a tiny notebook, which include a personal handwritten message for the receiver and 15 more empty pages for them to use as they wish.

Letterpress business cards
Each letterpress business card includes a personal handwritten message from Ecker

15. Bentply

Letterpress business cards
One minute it’s a business card, the next it’s a Gerald Summers armchair. Cool, huh?

What you see here is one of the neat letterpress business cards for furniture shop Bentply in Marylebone, London that can be made into mini bent plywood chairs.

The ingenious card design was created by art director, writer, and designer Richard C Evans and produced by Elegante Pressin Lithuania. Just follow the instructions and the kiss-cut card can be folded into a miniature of the iconic 1934 plywood armchair designed by Gerald Summers.

Creative business cards
All Elegante’s products are hand made using 100% cotton (tree-free) paper and oil based inks

16. Dot Design

letterpress business cards
The distinct shape makes these creative business cards instantly striking

Printed using two colours on both sides and die cut to shape, the carefully placed design elements on each side of these letterpress business cards avoid any overlapping. They’re printed on 425gsm cotton and are instantly striking. Creating such a distinct shape really makes these creative business cards one of our favourites.

17. Derek Welsh

letterpress business cards
The sleek finish represents Derek’s work perfectly

Glasgow Press achieved a monster 810gsm when creating these letterpress business cards by bonding 270gsm sheets of Ebony, Citrine and Bright White to achieve a stand out card for furniture maker Derek Welsh. Black letterpress to the Bright White side with the gloss black foil to the Ebony card really represent, making this a unique creation.

letterpress business cards
The gloss foil finish really makes these creative business cards stand out

18. Whitney Shaw

letterpress business cards
We love the colours used on this letterpress business card

The colours really won us over with this letterpress business card. Encorporating a variety of fonts, as well as a playful illustration, this is a design that would surely be cherished by the receiver. The use of larger ‘W’s’ also allow for Whitney’s name to stick in the mind, whilst keeping the design aspects to a minimal affect.

19. Bryon Darby

letterpress business cards
The clean-cut, minimal approach works for this letterpress business card

Often, less is more when it comes to design. It’s certainly the case for this letterpress business card for photographer Bryon Darby. Using a minimal colour scheme and using the letterpress to create a series of beautiful patterns, we love its simplicity.

20. For Luca

letterpress business cards
A lovely letterpress card for a worthy cause

When Luca was just three years old he was struck down with Meningococcal Septicaemia – losing both of his legs along with muscle and skin below his hips. When his dad contacted Blush Publishing for some business cards to help with Luca’s campaign they fired up the presses and produce this delightful design. We love the bright blue typography placed on a simple, white backdrop. Instantly eye-catching.

21. Amy Weibel

The typography is simple but effective

Amy Weibel is a art director & designer, with a passion for food, travel, technology and of course design. She is currently based in the bustling neighborhood of Chinatown, New York. The typography treatment on each side is simple and effective; whilst the trim is bold and memorable.

The trim is bold and memorable

22. Forge

The wraparound sticker is an inspired touch

A self-promotional business card for the design studio Forge. “We wanted a card that would communicate our hands-on and unique approach to design solutions,” they explain on their site. “We sourced the paper ourselves and used two printers.” The offset typography is pleasing, printed on a 200lb uncoated stock, with the wraparound sticker an inspired touch.

23. Kitty’s Beauty Parlour

The card was letterpress-printed on double thick 236lb cotton card stock

Print & Grain is full service graphic design and letterpress studio. They offer original custom design and letterpress printing. Print & Grain encourage a collaborative effort with their clients during the design process to create a unique and personal card, which will be letterpress printed on double thick 236lb cotton card stock, using an antique Kelsey Excelsior printing press. A strong example from their portfolio is this creation for Kitty’s Beauty Parlour.

The studio uses an antique Kelsey Excelsior printing press

24. Mei Yen Chua

The card combines slab-serif and serif typefaces

A variety of techniques were used on this business card for Mei Yen Chua. The experimental techniques using overlays of ink works wonders. Each shape also been embossed into the card, combining slab-serif and serif typefaces. A confident and eye-catching card.

25. Moglea

We love the incorporation of natural images into this card

Moglea is a boutique letterpress stationery design studio created by Meg Gleason. She loves bold pops of colour, hand-lettered typography and intricate floral and geometric patterns. The letterpress studio is within a farmhouse in Western Lowa, and this influences her style, which incorporates animals and nature into the designs.

26. Denim Geek

letterpress business cards
Blush pull off yet another beautiful letterpress printing job

This gorgeous letterpress business card comes from those talented folks at Blush – a bespoke and custom letterpress printing service that offers 
wedding stationery, personalised correspondence, greeting cards and of course, business cards.

Even though white letterpress ink is often extremely diffcult to work with and doesn’t give a very good solid on dark stock, it works extremely well in this design. The card was printed on 1000mic recycled board, so it’s good for the planet too!

27. Leigh Cameron

letterpress business cards
We love the stark contrast within the colour choices

The guys over at Typoretum have an incredibly impressive array of letterpress business card designs on offer. Specialising in comprehensive letterpress business card and stationery design, the printing service can work from artwork supplied or help you to design your own.

This creation for Leigh Cameron is simple, elegant and engaging. We love the stark contrast of the green coupled with the grey. The cards are printed on 750 micron recycled greyboard.

28. Britt Boyd

letterpress business cards
Combining stunning graphic design and typography, this is one of their best examples

Print and Grain are quickly becoming known for their letterpress business card offerings within the design industry. Based in Portland, Oregon, they also provide letterpress inspired greetings cards through their Etsy shop.

This design for Britt Boyd is a stunning feat in typography and graphic design. The beautiful graphics, combined with an eye-catching font and coloured edges, make it one of their best examples.

29. Ninja star

letterpress business cards
A letterpress business card that will get ‘the point’ across

Now, here’s a letterpress business card that’ll catch a potential client’s eye for all the right reasons. Showcased on Beast Pieces, this design will certainly get ‘the point’ across (sorry!) Although it might not be the ideal shape and size to slip into your wallet, the understated letterpress design and carefully placed typography make it a must in this list.

30. Nili Studios

letterpress business cards
Print & Grain triumph again with this letterpress design for Nili Studios

Another addition from the guys at Print & Grain, this lovely-looking letterpress business card was designed for creative agency Nili Studios. The nautical-inspired theme is perfectly executed with the chosen soft blues as well as the waves and anchor logo. The choice of font is also a brilliant choice, as its readability is showcased.

31. Druckerei Eisenhardt

letterpress business cards
The bold choice in size, shape and typography make this business card stand out

A printing press that’s been operating for over 42 years,Eisenhardt is a trusted German company offering the best in letterpress design. A family-run business, it works with both modern and traditional techniques in the centre of Frankfurt.

This letterpress business card design immediately caught our eye thanks to its bold choice in size, shape and typography. The orange-on-white colour scheme enables the font to stand out whilst the understated letterpress gives it the edge it needs.

32. Stu Horvath

letterpress business cards
Dolce Press showcase the intensity of one-colour printing

Dolce Press is a boutique print shop and design studio located on the eastern end of Long Island. Specialising in letterpress, the design studio creates one-of-a-kind prints.

This letterpress business card was printed on super-thick kraft stock in black ink, showcasing what can be accomplished with one-colour printing. The letterpress cards were finished off with black edge colouring and printed on a Chandler & Price 8×12.

33. Silk Lab

letterpress business cards
We love Lettera Magica’s spin on the usual rectangular shape

A workshop based in Cracow, Lettera Magica combined their love for tradition, perfection, patience and respect for nature to create some of the most stunning letterpress business cards we’ve ever seen.

This design was crafted for creative studio Skill Lab and was printed on Duplex Orange paper. It measures at 90 x 55mm, which makes it the perfect size to slip into your wallet. We love their slight change in the usual rectangular offering.

34. J. Fletcher Design

letterpress business cards
J. Fletcher perfectly echoes his client’s ethos, with this natural-looking offering

J. Fletcher has been producing graphic solutions for clients ever since 2004. Describing himself as a ‘one-man show, not a one-man band’ his executions in branding are faultless. This design – created for Charleston Naturally – perfectly captures their ethos.

They’re made from the highest quality products – often organic, mindfully local. Fletcher echoes this ‘natural’ ethos with this beautiful letterpress business card design.

35. Jason Turiff

letterpress business cards
The ink used for Jason’s business card was the perfect balance of blue and green

Another addition from Dolce Press; this time, for studio engineer Jason Turiff. Every colour here is hand mixed from a base of 16 colors which allows Dolce Press to achieve the full spectrum of Pantone colours and more. The ink used for Jason’s business cards was the perfect balance of blue and green. They were then printed on Crane’s Lettra (Pearl White, 300gsm) paper.

36. Fizz

Letterpress printing
These coasters complement Fizz Creative’s ‘Design straight up with a twist’ motto perfectly

Graphic designers Jasen Melnick and Katie Major used letterpress printing to create these awesome retro-style coasters to promote their award-winning boutique studio Fizz Creative. Their letterpress business cards design perfectly complements the duo’s motto, ‘Design straight up with a twist’.

Featuring two Pantone colours, and printed on thick, ivory white stock, these business cards won awards from the AIGA Cleveland Design Competition and from the HOW+Print Color in Design Competition.

37. Dare

Letterpress printing
These letterpress printed cards feature ASCII images of individual employees. We want some!

We love these innovative designs by creative agency Dare. The company describes itself as ‘a merger of a traditional agency and digital shop’, so it wanted letterpress business cards that also combine old with new. Employees were invited to upload a picture of themselves and enter their email and phone numbers. An ASCII image was then automatically created using characters from their contact details.

The company then chose to letterpress these images onto individual business cards in order to re-emphasise the idea of traditional techniques meeting digital. Awesome.

38. Pablo Abad

Letterpress printing
Pablo Abad created this sans-serif typeface specifically for this project

This beautiful design was created by Madrid-based graphic designer and art director Pablo Abad. In a quest to create a new visual identity, Abad designed a bespoke geometric sans-serif typeface gara.

The clean, yet edgy letterpress business cards design makes it eye-catching but still easy to understand. Abad’s love of illustration and typography is clear in this very personal project.

39. Shyama Golden

Letterpress printing
Shyama Golden – ‘It Rhymes with Llama’. As in Shama-Llama, get it?

It’s always an awkward moment when you can’t pronounce someone’s name or get it wrong. To ensure that never happens again, artist and designer Shyama Golden designed these funky letterpressed business cards with her catchphrase ‘It Rhymes with Llama’.

Printed on Cranes lettra cotton paper, with 100% hand-drawn type and a hand-drawn llama, these beauties are one-colour letterpress printing at its best.

40. Luke Lucas

Letterpress printing
Luke Lucas’s card uses specialist ink which makes it glow in the dark. Ooooooooh!

After developing a new script-based logo, typographer and graphic designer Luke Lucas decided to update his business cards to show it off. Featuring his name on one side and his logo on the other, Lucas’s card is no ordinary one.

Printed on 600gsm Lettra cotton stock, this letterpress business cards design uses phosphorescent ink so that in regular light it appears as a simple debossing but when the lights dim, the logo illuminates in a bright green. Very cool.

41. James Prunean

Letterpress printing
James Prunean lovingly created these awesome business cards for his brother Ovi

As is often the case, simplicity can work wonders. A perfect example of this is these gorgeous letterpressed business cards created by graphic designer James Prunean for his brother, Ovi.

Prunean used strong, bold colours that immediately catch your eye and at a first glance look very similar to Pantone colour cards. Printed on 220lb cotton paper, the soft yet strong impression has reportedly helped painter Ovi beat off a lot competition. And we can see why!

42. Jennifer Daniel

Letterpress printing
Jennifer Daniel loves unicorns. Can you tell?

These quirky but cool business cards belong to Bloomberg Buisnessweek’s graphic director Jennifer Daniel. An internationally recognised illustrator, designer and art director, Daniel is also a lover of unicorns.

After having a hard time choosing from her designs, she decided to have three printed, each one as memorable as the next and featuring her RSI-inducing contact email along the bottom. We wonder whether she’s got any unicorn grooming business off the back of them?

43. Plaid Lab

Letterpress printing
Does anyone else want to stroke this business card?

The Plaid Lab is the portfolio of designer Mark Saunders, and the design for his business card was clearly inspired by that name. The gorgeous, brightly coloured letterpress business cards design features a plaid pattern, consisting of crossed horizontal and vertical bands in two or more colours.

The cards were letterpress printed on heavy 220lb Cover Crane Lettra Pearl White stock and used overprinting inks to create the rich and saturated colours. The printing is so good, it actually looks like Plaid material forms part of the card. An obvious but very clever design.

44. Rabbit Hole

Letterpress printing
With this business card you also get a cute desktop companion

This beautiful design grabbed our attention as soon as we saw it. Why? Not only because of the gorgeous impressions in fluourescent green with a double hit of silver on Beer Matt Board 390gsm but also because you can turn it into a rabbit.

Yes, you read that right. Several shapes are die-cut out and can be put together to create a stylised bunny. Cool, huh? We bet the guys at letterpress and design agency The Hungry Workshop had great fun creating these for The Rabbit Hole Ideation Cafe.

45. Still Liquor

Letterpress printing
The message here is not to drink and drive. Hence the word still.

This classy design was created for Still Liquor, a bar and micro distillery in a reclaimed bootleg liquor and auto shop in Seattle. To convey the bar’s rich heritage, New York-based multi-disciplinary design practice Javas Lehn studio decided to use the iconic Ford Model T silhouette, bold type and unexpected contemporary compositions.

The final, clean design was then letterpress-printed onto Eska board stock, using a bright shade of red to accentuate the bold type. If the bar itself is a classy as its business card, we’d definitely drink there.

Have you seen a great example of letterpress business cards? Let us know in the comments!

WordsSammy Maine

Sammy Maine is a reporter at Creative Bloq.

Meet Bossy,

A Cute Desktop Assistant That Wants You To Be A Better Worker

The device is concept designed to show a future where our electronics can help us organize our work lives beyond just installing more and more productivity apps.

It might seem like working long hours would make you better at your job, but the opposite is actually true: It can make you less productive and lower the quality of your work.

(Even worse, it also makes you more likely to die early).

A new product concept from U.K. that design student Lucas Neumann aims to help by telling you exactly what to do so you can finish work early–like a friendly version of a robot boss.

Neumann designed the product for freelance workers with flexible hours, who might be more likely to procrastinate and get distracted by an avalanche of information online.

But it probably could be helpful for anyone who’s stuck in front of a computer all day and has trouble concentrating.

The tool, called Bossy, is a physical device that sits on a desktop, rather than yet another app. “When I was doing research, I started installing all of these organization apps in my phone and on my Mac,” says Neumann. “But I realized that after time we stop engaging with particular apps because we’re dealing with too much stuff.”

Bossy is a constant reminder of whatever’s next on your to-do list, so you never have to open an app or check in.

“I realized that if you have something that sits there with you all the time, off your phone and computer–while at the same time connected with everything–it might be easier to create a relationship with it and a long-lasting kind of engagement,” Neumann explains.

The simple white device connects to your calendars, to-do lists, apps, and wearable tech, and then displays your top three priorities.

Once you finish something, you push a button–made squishy so it’s more satisfying–and the device records your progress on a separate desktop and mobile app, offering badges and other rewards as you improve.

Over time, Bossy learns how you work in order to be more helpful.

It can also remind you to stand up, stretch, drink water, or take breaks, and if you’re feeling overwhelmed by various feeds online, a swipe of a button on the device will shut down the noise.

Neumann created the design for an RSA student challenge that asked for solutions for the way we’ll work in the future. “It’s looking 15 or 20 years ahead,” he says. Still, the technology and predictive analysis used in the design are already available, so in theory, the product could be built now.

For the moment, Neumann wants to tweak some details of the hardware–a special white screen that he wants to use, for example, isn’t available yet on the market.

But as he researches the concept with freelancers at coworking spaces and elsewhere, he’s already getting requests to make it right away.

ADELE PETERS is a writer who focuses on sustainability and design and lives in Oakland, California. She’s worked with GOOD, BioLite, and the Sustainable Products and Solutions program at UC Berkeley. Continued

When he is not teaching samba classes and helping kids in rural Brazil with their homework, Lucas Neumann de Antonio designs. And when he does, it ends up on FastCo.

 

 

Distributed Systems and the End of the API

This is a written (expanded) narrative of the content from a talk I first gave at PhillyETE on April 23rd, 2014.  (selected as one of top posts)

It mostly follows the flow of the presentation given then, but with a level of detail that I hope enhances clarity of the ideas therein.

The talk’s original slides are available, though the key illustrations and bullet points contained therein are replicated (and somewhat enhanced) below. When audio/video of the talk is published, I will update this page to link to it.

I have two claims of which I would like to convince you today:

  1. The notion of the networked application API is an unsalvageable anachronism that fails to account for the necessary complexities of distributed systems.
  2. There exist a set of formalisms that do account for these complexities, but which are effectively absent from modern programming practice.

The outline of this piece is as follows:

Let’s start with some definitions, just because “distributed systems” and “API” are broad terms that have many connotations to different people. I’m talking about a particular (large) subset of the many different aspects of these things.

Distributed Systems

There is a tongue-in-cheek saying that captures this generalization and defines what a distributed system is:

A distributed system is one where a machine I’ve never heard of can cause my program to fail.

I don’t know who first coined that characterization, but it is something of a meme among those that study and work within fields related to distributed systems.  There are many formal definitions of “distributed system” all over the internet, but my personal riff is:

A distributed system is one that is comprised of multiple processes that must communicate to perform work.

The bottom line is that, given the ambient nature of the networks that surround us and the dependence we have upon those networks for so many of the tasks our programs, clients, customers, and users take for granted, nearly every system we build is a distributed system.

Unless your software runs in a totally isolated environment — e.g. on an air-gapped computer — you are building a distributed system.

This is problematic in that distributed systems exhibit a set of uniformly unintuitive behaviours related to causality, consistency, and availability.

These behaviours are largely emergent, and spring from the equally unintuitive semantics of the non-locality of the parts of those distributed systems and the networks that connect them.

None of these behaviours or semantics are related at all to those which we — as programmers and engineers — are typically trained and acclimated to expect and reason about.

Note that even if you are doing something small, or “normal”, or common, you are not immune to these challenges. Even the most vanilla web application is definitionally a distributed system.

By sending data from one computer (e.g. a server) to another (e.g. your customer’s web browser), you end up having to contemplate and address all sorts of problems that simply don’t exist when you run a program in a single process on a single machine that doesn’t touch the network: consistency, coping with non-availability (i.e. latency, services being down, timing-related bugs caused by long-running computations or things as simple as garbage collection), dealing with repeated messages from clients with spotty connections, and more.

If you’ve not been bitten by these things, that is evidence of luck (or, of your not having noticed the problems yet!), not of your being immune, or otherwise that what you’ve built is somehow not a distributed system and so isn’t subject to these challenges.

Application Programming Interface (API)

Again, there are lots of definitions of the “API” term; this is mine:

An API is the set of names we interact with in our programming languages and libraries.

This concept — naming the data, types, objects, methods, modules, namespaces, and so on in our programming practice — has been around almost as long as people have been programming computers. More concretely, its origins lie in the context of imperative programming languages and libraries. Common examples include the standard C library, Win32, POSIX, and all the names of things in your favourite Javadoc, Rubydoc, library or function man pages, and so on.

When people started hooking up computers over networks, it was natural to want to carry along this notion of using language as a way of naming things we interact with programmatically.  Of course, assigning names is not an issue; doing so is essential to being able to talk about them at all. The problem is that APIs are fundamentally only nominal descriptions. We assign names to the data and operations and objects our programs manipulate, but there is nothing in such a shorthand that talks about the semantics or limits or capabilities of those things. To abuse a Perlisism, the name of a thing is a perfect vehicle for hiding information.

Thus, when we are doing some network programming:

client.send(data);

There is nothing about the API we’re touching that informs us of the semantics of the operation we’re attempting to perform. What are the bounds on how long it will take? What are its preconditions? What it should do (or what we should do additionally) in the case of a timeout, or failure, or partial success? The undefinededness is unlimited. While the lack of clearly identifiable, defined semantics is problematic even in the case of non-networked, non-distributed, single-process context programming, it is near-fatal when you are working on a networked, distributed system: you must take on the job of identifying and understanding those semantics yourself, and situating your use of APIs such that they account for the failure modes implied by the network and other aspects of the distributed system you’re building.

The API Problem

So, using names as the sole source of conceptual leverage while programming is problematic, and particularly so when it comes to network-related APIs. This practice is likely so deeply woven into how we have cast the instruction of computers as a language problem that any change would be a monumental conceptual shift, far beyond what I want to address here. Rather, I’d like to take a look at other aspects of APIs in the distributed systems context that I believe are fatal.

Here is a rough, temporal progression of the major network API technologies that have seen wide use:

RPC → #{DCOM, CORBA} → RMI → XML-RPC → SOAP → REST → #{‘REST’, Thrift}

As far as addressing the problems of distributed systems go, these are all fundamentally equivalent. It’s possible to quibble about this on the edges, insofar as e.g. RPC mechanisms obviously have particular implementation details that are not shared by the large umbrella of HTTP APIs (which I denote above as “‘REST’”), but I claim that they all share and provide the same set of fundamental semantics:

  • a request/response lifecycle
  • always synchronous
  • presume a point-to-point communication topology (i.e. two party, client/server communication)
  • operations provided by these mechanisms are nearly always imperative, implying mutable data models and side-effecting operations
  • few constraints placed on acceptable data models or representations are imposed

These characteristics are shared by the programming language heritage that originally defined the notion of “an API”: one caller invoking operations synchronously on one callee, providing a request (arguments) in exchange for a response (a return value, sometimes), where the data involved can effectively be anything — very often mutable structures that are modified in-place — all while causing some side effect.  Unfortunately, these characteristics are a large part of why APIs are a fatally bad mismatch for the job of supporting the communications between actors in a distributed system, which are frequently not best characterized as request/response, more commonly asynchronous than not, are often not ideally point-to-point, and which suffer as much or more from mutable data models and side-effecting operations as “regular”, single-process, non-networked programs.

(At this point, you might object, as some of the API technologies I called out above do provide limited support for e.g. asynchrony.  For example, an HTTP API could reasonably respond to a request that will be processed asynchronously with a 202 Accepted status, plus a URL of a “status” resource that would eventually redirect to the final result of processing the original request. However, allowances like this amount to nothing more than idioms and “best practices”, not substantive solutions to the limitations described above. Implying otherwise would be analogous to claiming that a less capable programming language A might be argued to be equivalent to a more capable language B, simply by dint of both being Turing complete.)

Tweet text: "People hear 'RPC', and giggle, smugly shaking their head while pounding out REST integrations."

APIs: Sisyphean programmer convenience

While APIs provide a particularly narrow portal to the outside world, they have done wonders for programmer convenience. Here I am talking about one particular convenience that programmers implicitly value very highly when working with network APIs, and that’s the maintenance of an isomorphism between function or method calls as found in typical imperative programming languages, and the various manifestations of network APIs, regardless of the particular technology in use.

For example, here are three common ways one might see usage of a single “create” API call exemplified:

api.create(arg1, arg2);
[arg1, arg2]
[arg1, arg2]

The first could be either an in-process method call that doesn’t touch the network, or an RPC call that uses programming language-native stubs or a client library and which implicitly reaches across a network. The second is typical of a strictly imperative HTTP API construction. The third is the “proper” REST corollary that uses URIs, HTTP verbs, and resource representations in particular ways to achieve the specific semantics defined by REST.

In each of these cases, nominal concerns are paramount, but network and other operational semantics, failure modes, notions of causality and consistency are entirely unaccounted-for by the different API mechanisms. These unstated costs are implicitly retained by the programmer that happily uses these mechanisms, and must be balanced by either manual accommodations, error handling, and deep considerations of application-specific causality and consistency invariants…or user-visible and business-tangible failures.

One particular irony is that, while more modern API technologies have been built with full knowledge of the accepted failings of RPC, many providers of e.g. HTTP APIs provide and specifically encourage programmers to use “client libraries” for one’s preferred language, effectively restoring the classic RPC programming experience, where it’s difficult to impossible know which calls will result in strictly local computation, and which will incur one or many network operations and all of the complications that that entails.

stripe-http-api

It is as if we are supposed to carry on calling what look like regular functions and methods, and that everything will just work, like magic!

This is convenient, and feels very familiar, but does not address the problem space that we are working in.

The API: an anachronism

“The API” as the fundamental point of integration between parts of a distributed system is an anachronism, a hold-over from other, simpler programming contexts that predate “distributed systems” as a discrete concept:

  • APIs necessitate an intense coupling between actors in many ways, principally:
    • By admitting only two-party client/server architectures, despite the actual myriad of application and network topologies that exist; anything else needs to be constructed out of this point-to-point primitive, or pushed into silos that do provide for more complex topologies (e.g. queues, system busses, etc).
    • By allowing arbitrary data representations to be used: in order to talk to your API, I need to reconcile how my client represents data with what your server expects and produces. It is impossible to later talk to another equivalent API without repeating this process.1
  • Common computational tasks necessitate asynchrony.  Though some patterns and common workarounds exist, APIs are fundamentally always synchronous.
  • APIs, as a class of technology, disavow the fundamental complexities of distributed systems:
    • Coping with network failure modes (latency, disconnection, offline contexts)
    • Making consistency choices relevant to our applications, and being aware of the choices being made for us by our underlying technology
    • Being aware of the impact that our consistency choices have on availability
    • How our data models and data representations influence and sometimes determine what is and is not possible in terms of concurrent activity by different actors

To be precise, there are seven basic network (and therefore, system) topologies, and an infinite number of permutations combining them:

500px-NetworkTopologies.svgWhen you consider the systems you build, you’ll recognize these topologies within whatever diagrams you draw. Meanwhile, APIs provide no way to naturally talk about anything other than two-party client/server communication. Just as any limited programming language can lean on Turing completeness to claim capability, we can lean on APIs and say that we can build any topology we like with them; while factually correct, this is an flaw, not a feature.

Acknowledge the network or fail

What distinguishes a network API from a “regular”, in-process, single-node API is…the network. That’s a tautology, but one worth making given the apparent primal desire of programmers to ignore the network in their modern practice (viz. RPC and client libraries for network APIs that obfuscate their true nature).

When you try to get computers to work across a network, the network is an integral part of the resulting larger system. If you do not account for its inherent nature, your larger system will fail along with that network, in ways large and small. The first step in addressing this is to understand the ways in which networks can fail, and how those failures present to your programs.

I’m by no means an expert in network failure modes (keeping up with Kyle Kingsbury’s work with Jepsen is a good first step, if you are coming from a programming-with-databases background and are looking for something accessible), but the basic network failure modes include:

  • Partitions
  • All the vibrant colours of latency
    • Variable latency, which can be caused by everything from overloaded switches to garbage collection stopping the world on a remote machine
    • Complete and permanent loss of interconnect
    • Offline operation
  • Reordered messages
  • Repeated messages

Many of these types of failures have been most well-discussed in the context of distributed databases, but they apply just as forcefully to the entirety of your systems. A partition among your application servers can cause clustered, in-memory sessions to diverge, leading to invariants being violated both server-side and in users’ web browsers as they are fed data from different divergent copies of their session. Increased latency (whatever the cause) can trip timeouts between different parts of a system, leading to spiking reconnect attempts and thus cascading, catastrophic latency. When a machine becomes totally unreachable, what parts of your system can carry on, and can the data it was serving be recovered elsewhere? What are your users’ expectations when they have no connectivity at all? When a client resends a message because it thought your service didn’t receive its first attempt due to a timeout (except it did), what do you do?

In short:

  • Your network’s problems are your system’s problems. The corollary of this is:
  • My network’s problems are your system’s problems.

That is, the networks used to reach the people and devices and vendors at the edges of your system are almost never owned and operated by you, yet your system is subject to their failures as well.

Different technologies and different data models will provide greater or lesser inherent protection from the capriciousness of networks. Of course, network API mechanisms provide exactly none, and the data representations that are commonly shipped over those mechanisms (e.g. JSON, XML, YAML, plain text, and so on) are equally of no help.

Consistency decisions affect everything

Consistency and consensus are huge topics, and so my treatment of them here will be wanting. Much like networks and their failure modes, you cannot design or build a robust distributed system without being aware of them. However, unlike networks — the problems of which are effectively a force of nature — you have within your power the ability to choose what your system’s consistency guarantees will be, presumably based on the needs of your customers, users, and organization.choose-wiselyThere are many types of consistency, and there’s no way that I could enumerate all or even some of them. Further, there is not a spectrum of consistency; the set of available options forms something more like a multidimensional space, as there are a lot of considerations that go into choosing a set of consistency guarantees. But, the most common cases as far as I can tell include:

  • Strict linearizability, where all actors in a system synchronously acknowledge each write to any shared data, yielding a global total order of all operations. Systems with this characteristic act as if their entire state is held within a single atomic reference, as found in many programming languages with such a concurrency primitive. This is incredibly expensive in terms of computational and communication overhead, but perhaps corresponds most closely with our intuitions about what happens when, and where.
  • Causal consistency, where logically temporal relationships between dependent changes to shared data are tracked, yielding a partial order of all operations. Causal consistency requires much less consensus overhead for any given request to proceed (compared to strict linearizability), but ensures many desirable properties within a system with such a guarantee. One of these is the ability to read your own writes; without some mechanism to enforce causal consistency, it is possible for e.g. a web client to write a value through an HTTP API, and then see the previous value when it performs a read or query some time afterwards.
  • Eventual consistency, where concurrent writes converge (perhaps with conflicts) such that different readers will all eventually see the same result at some point in the future when all prior writes have been applied. This implies no order at all to operations within a system; in effect, the only guarantee of eventual consistency is that of liveness, a term used in distributed systems literature to imply that all actors within a system will continue to propagate writes until all actors have seen every write.

The choices you make with regard to consistency have a compensating impact on the availability characteristics of your system. This relationship is known more formally as the CAP theorem (as introduced, and then proven), but the basic dynamic of the tradeoff should be obvious given a couple of moments thinking about the relationship between consistency and availability:

  • Insofar as every actor in your system has to synchronously acknowledge each write made by others, then those actors (a.k.a. servers, or services) cannot accept any new work until that acknowledgement process completes. During this time, your system will appear to be down (unavailable) to external parties.
  • If no consensus is required among different actors (as in the various eventual consistency models), then every actor can always accept new work. Barring resource exhaustion or exigent factors (such as a network lapse), your system will always be up (available) and responsive to external parties.

Note that different parts of a system may have different consensus requirements (and thus its consistency guarantees), and those requirements and guarantees may change over time (potentially rapidly and regularly in order to accommodate user expectations and coherence demands of different kinds of data).

Just as with networks, different technologies and data models provide more or less leverage and flexibility in accommodating or enabling different degrees of consensus, consistency, and thus availability. Unsurprisingly, network API mechanisms again provide exactly nothing: they presume nothing, except that an open socket is available for use.  It is up to you to consider these issues properly, and either prevent or resolve any inconsistencies within your system.

Likewise, while plain text does offer some meagre ways to address concurrent inconsistent changes (e.g. two actors modifying the same text document concurrently can often be resolved sanely via a diff mechanism), the “richer” data representations that are favoured by most API services and clients (again, JSON, XML, etc) are fundamentally opaque and in general make reconciling independent changes impossible in a consistent way without special, often domain-specific intervention. Further, typical implementations of data structures in our programming languages provide no ways to represent or reconcile concurrent changes at all. Together, this means that concurrent actors moving state around or representing operations using these data structures and representations have no generally applicable way of resolving conflicting concurrent changes. This forces programmers to regularly re-implement such resolution mechanisms; or, more commonly, rely completely upon centralized backend databases to allow concurrency and enforce consistency, semantics that become less reliable further away from those centralized authorities.

What do we want?

Okay, I’ve just rambled on for a long time about how distributed systems have some challenges above and beyond “regular” programming, and that APIs as we know them today do not meet those challenges and may even exacerbate them by nudging us towards glossing over their existence and scope. It’s a pretty bleak picture, and you may walk away at this point and consider me a crank.

I’m at peace with that, but I defy anyone to claim that programming networked services and applications is easy. Or that the interactions between e.g. disparate databases, application servers, caching services, load balancers, client browsers, a couple of message queues, four vendor services APIs, and that nutty Hadoop job are understandable. Or that errors stemming from the incidental complexity of the technologies involved in building such a system are rare. Or that when something does go wrong in that easy-to-build, understandable, rarely-failing distributed system, it’s easy to diagnose the problem.

None of these things are true. Of course, I’m not saying that building distributed systems will ever be as easy or as simple as other types of programming and engineering. Further, I’m not even saying that APIs are themselves the root of all of this difficulty and complexity (though they play their part); APIs just happen to be a convenient, familiar, and obvious pain point that exemplifies the primitive nature of the raw materials we’ve imported from other programming contexts to build distributed systems.

Building and reasoning about distributed systems should be easier and simpler than it is today.2 I’d like to suggest that, in order to achieve this, we need different tools than the ones we have at our disposal now.

Let’s step back. What do we want? In any distributed system we build, we want two things:

  • Communication — the ability to share data among the various actors participating in our system
  • Computation — the ability to consume and transform that data, producing new data as a result that is perhaps itself communicated

Everything else about building distributed systems is incidental. We need building blocks that:

  • allow us to control and apply these capabilities however our particular needs demand
  • actively prevent us from committing obvious errors in reasoning about the irreducible complexity of distributed systems
  • even better, are architected such that committing such errors are fundamentally impossible

That’s a tall order, and not one that can be fulfilled completely; hell, we haven’t even fulfilled one of them — computation — in the much simpler context of non-distributed, non-networked programming. But, when looking for better tools to work with, we need to keep looking towards these two primitive capabilities, lest we steer elsewhere.

We’ve been here before

The history of programming languages has seen many analogous transitions.

Decades ago, writing software required working with extremely low-level programming languages, such as assembly and C. Relative to the higher-level languages developed since, those languages are very difficult to use effectively, have been and continue to be associated with much higher rates of error, and often prevent even contemplating building software systems approaching scales that are commonplace today. I think the same kind of progression can and must happen within the context of designing and building distributed systems: we simply cannot keep dragging along this notion that one can continue to work at or near the level of spitting data out of sockets and expect to build robust and understandable distributed systems on top of such primitive primitives.

Put more simply, going back to picking on APIs, what will complete this analogy?

assembly/C : Java/Python/Clojure :: APIs : ???

Just as our predecessors identified problems with machine code and assembly and constructed abstractions in higher-level languages, we must rise above the metal of sockets, RPC, APIs, and so on.  I believe that much of this work is in identifying what not to do and what not to offer, just as higher-level programming languages are largely characterized by their constraints compared to what came before. In both contexts, what we find to be confusing and error prone — e.g. userland goto, manual memory allocation, and in-place mutation of objects in the case of programming languages, and things like accommodating network failure modes and implementing appropriate consensus mechanisms and consistency levels in distributed systems — are exactly the things that we need to abstract away from in day-to-day concerns in order to make progress.

My appeal to authority

Thankfully, many people smarter than I have been thinking about these sorts of problems in the context of distributed systems for some time. Perhaps not as long as people have been thinking about programming language problems, but close: soon after networks “happened”, people became aware of the difficulties of having multiple computers agree on things and compute manipulations over the data that they share.

Leslie Lamport

In many ways, Leslie Lamport’s 1978 paper Time, Clocks, and the Ordering of Events in a Distributed System was the starting shot for serious, principled research into the challenges of distributed systems. (And the beginning of a body of work which contributed significantly to Mr. Lamport’s receipt of the Turing Award earlier this year.)

What I appreciate most about this paper is that he talks about the difficulties of communication among concurrent, distributed actors as a physics problem. Contemplating the challenges of a distributed system reveals this as more than a convenient analogy, and closer to actual fact: the essential qualities of the distance between separate actors and the progression of time that paces their communications define what is and is not possible.

Mr. Lamport talks at some length about this perspective in a conversation he had with Erik Meijer (@ 15m30s):

Most people view concurrency as a programming problem or a language problem. I regard it as a physics problem.

Mr. Lamport’s work and this inclination to reach for provable approaches to problems grounds modern research in distributed systems. While much of what we do in software might be charitably characterized as engineering (nevermind Science, or even “science”), there is some comfort to be had that a thoughtful approach to distributed systems problems should flow from a more reliable basis than even the most well-designed library, framework, or language. The emphasis being on provable formalisms over clever engineering and implementation details gives me hope that solutions to these problems, once found, may be as reliable as the mathematics and understanding of physics upon which they are built.

Sound approaches nearing practicality

Over the last ten years or so, a number of sound approaches to some of the fundamental problems of distributed systems have been developed. I’ll focus on two strongly-related threads of progress:

Both of these approaches constrain the types of operations that your system can perform in order to ensure convergence over time of changes to data shared by uncoordinated, concurrent actors, and to eliminate network failure modes as a source of error. Achieving this vastly simplifies the challenge of building robust distributed systems, just as certain advances in language design simplified the problems of programming by constraining what we could express in our programs.

As I foreshadowed earlier, what’s essential about both of these approaches is not novel engineering, but the provable formalisms upon which they are premised: CALM has its roots in temporal logic, while CRDTs depend upon the algebra of semilattices.  You should understand these formalisms, at least in passing, if you are to leverage their concrete implementations with confidence and to good effect.

(Before you think that this is all in the realm of purely academic research, consider that, for example, CRDTs are already in production, and are included as a banner feature in the impending next release of Riak. These are not strictly high-minded proposals, but actionable progress in developing an understandable basis for building robust distributed systems.)

I happen to be far more familiar with CRDTs and semilattices than I am with CALM’s temporal logic, and so that will be my focus for the next section. (Much of it is also true for CALM and implementations of it.) The good news is that the math in question is really quite easy: really little more than secondary school-level algebra.

(If you’re reading this prior to May 15th, 2014, you may be interested in a talk I am delivering on that date at Papers We Love in New York discussing CRDTs using the paper that originated the ‘CRDT’ term as a vehicle.)

(Bounded) join semilattices

A lattice is a partially-ordered set where, for any two members of the set, there exists both a least upper bound (a value that is “greater than” both members, however that relation is defined for the types in question) and a greatest lower bound (a value that is “less than” both members). These relations are called the join and the meet, and if a partially-ordered set admits only one of them (i.e. it is partially-ordered in only one direction), then it is a semilattice. Finally, if a lattice has an absolute maximum or absolute minimum (often called top and bottom), it is said to be bounded.

CRDTs are premised upon join semilattices: partially-ordered sets of values for which a “greater than” relation holds for every member in the set. (Join and meet semilattices are mathematically equivalent; I believe it is largely just a matter of convention that join semilattices are used to describe and illustrate CRDTs, but the choice is not essential to their construction.) Finally, many CRDTs (based on those I am most aware of from the literature) are bounded.

a set semilattice

Contemplating bounded-join semilattices is reasonably easy once you have seen a couple of the simplest examples.  One would be the set of natural numbers, where the join operation is max. Another common example is a set semilattice, pictured here, where the join operation is set union.

This depiction of semilattices is very common, perhaps because it illustrates both the set-theoretic effect of the join operation (e.g. taking the maximum of two natural numbers yields the higher of the two, and taking the union of two sets yields a set containing all of the members of both sets), and hints at the intuition you should develop about how a semilattice shared among multiple parties might progress upwards through its possible states over time (logically and otherwise). A set semilattice with a join relation of union will never lose information: the value of each member will always increase in size as joins are performed over time, across participants sharing the semilattice. Said another way, this is critical within the context of distributed systems because it means that you can have concurrent modifications being made to a structure like this that is logically shared by many different actors, and those modifications will always yield the same value when joined.

Algebraically, join and meet operations for any semilattice must satisfy three axiomatic properties (shown here with algebraic as well as more programming-oriented formalisms):

  • Associativity, the ability to batch inputs to an operation in any way without affecting the result
    (a \cdot \!\, b) \cdot \!\, c = a \cdot \!\, (b \cdot \!\, c)              f(f(a, b), c) == f(a, f(b, c))
  • Commutativity, the ability to change the order of inputs to an operation without affecting the result
    a \cdot \!\, b = b \cdot \!\, a                         f(a, b) == f(b, a)
  • Idempotence, the ability to apply an operation multiple times without affecting the result
    a \cdot \!\, a = a                            f(f(a)) == f(a)

These are really useful characteristics even in our single-node, single-process programs that, when leveraged, allow us to reason much more easily and accurately about our programs’ behaviour in the face of concurrency and parallelism.  They provide treble benefits in a distributed context though. As long as the primitive operations over an implemented data structure maintain these invariants, it is a semilattice, and is sheltered from some of the most problematic network failure modes: commutativity ensures that reordering of network messages has no ill effect, and idempotence does the same if network messages are repeated.

Further, as long as liveness is preserved (the most minimal guarantee of eventual consistency), then semilattice semantics ensure that convergence of modifications and operations initiated concurrently occurs without conflict. Consider: there is no way in which set union (the join operation of a set semilattice) could not reconcile sets modified concurrently by different parties. The result will simply be the union of those sets, and always will be.

The astute observer might point out that not all useful operations are associative, commutative, and idempotent; removing items from a set is one easy example, but anything that might be characterized as causing a loss of information over time (removals, deletions, subtractions, etc) would qualify. The key here is that only the primitive operations over semilattices and CRDTs must satisfy those axioms; other operations can be implemented in terms of those primitives, thus yielding the key characteristics we seek. The good news is that there are good ways to do this. The details of how this is done I’ll leave for another day, or for you to discover.

Data models are everything

Despite the algebraic invariants that semilattices dictate, CRDTs can and have been built representing a wide array of data types, including:

  • counters
  • registers
  • sets
  • (multi)maps
  • dense and sparse lists/vectors
  • partially-ordered sequences
  • trees
  • graphs

This looks like a totally reasonable set of data types, all of which you would expect to have available to you in any modern programming environment. Recall though that all of these data types are Replicated (the ‘R’ in “CRDT”): they presume a network transport (with certain characteristics that I won’t detail here) that will push state or changes to it from one actor to another.

This means that, when using CRDTs to tie your system together, you don’t need to resort to using impoverished representations that simply never come anywhere near the representational power of the data structures you use in your programs at runtime. Instead of working out a way to distill your model data into JSON or XML so it can be shared with other actors in your system, you simply add the data in question to a CRDT, and let its replication implementation carry it abroad, with exactly zero loss of representational fidelity.

A further bit of good news for those of us that appreciate functional programming is that semilattices naturally encourage immutability. Insofar as each element within the set that forms a semilattice always “grows” relative to its predecessors and inputs, the most naive approach to implementing CRDTs (and a reasonable mental model regardless of actual implementations) is to maintain an immutable log of operations being applied or state being added. This means things like histories, rollbacks, and consistent snapshot — features and operations that are typically considered “advanced” today because of the incidental complexity of implementing them using the data and communication substrates that are generally in use — all come effectively for free.

Finally, let’s consider again the API, our original subject of ire. Use cases that might otherwise be addressed via an API can be entirely supersetted by using a CRDT. The transformation is a simple one: change imperative, side-effecting calls like this one:

api.setName(personId, "Chas");

into reified data that you add to a CRDT, which is thus replicated from a “client” to other actors (perhaps a single “server”, if you so choose):

{:person-id person-id :name "Chas"}

The change is subtle, but has a tectonic effects. “Operations”, when cast as data, become computable: you can copy them, route them, reorder them freely, manipulate them and apply programs to them, at any level of your system. People familiar with message queues will think this is very natural: after all, producers don’t invoke operations on or connect directly to the consumers of a queue. Rather, the whole point of a queue is to decouple producer and consumer, so “operations” are characterized as messages, and thus become as pliable as any other data. All of the same leverage applies when you replace APIs with CRDTs.

Pick a programming model

The great thing about CRDTs is that they do not require you to shift your programming practice wholesale. CRDT implementations have generally materialized as libraries, not specialized runtimes or languages, so you can readily use whatever programming language you like with a CRDT. I personally think this is a very good thing, as different languages and runtimes offer different capabilities and excel in different domains. While we’re talking about distributed systems here, much of the computation that we want to do is essentially local even when the inputs and results may be pulled from and pushed to a shared medium. Thus, having the freedom to choose particular programming technologies and tie them together via CRDTs feels like having the best of both worlds.

On the other hand, there are certain approaches to distributed systems that, in attempting to address their unique challenges, propose a total and holistic shift in programming practice. Bloom — the programming language that has been used to research and prototype the CALM theorem — is an effort in this direction. The vast majority of data structures that you manipulate as a matter of course in Bloom are lattices or other monotonic constructs. The tradeoff here is that while Bloom can statically identify which parts of your program are non-monotonic — in join semilattice terms, where your program state descends in the partial order, and thus is exposed to e.g. common network failure modes — Bloom is the only place where you can reap the benefits of that analysis.

Languages aside, I think there are a number of programming models that are particularly well-suited to the sorts of computation and domains to which the largest distributed systems are often applied:

  • Event sourcing
  • Stream-based computation (e.g. Storm)
  • I’m particularly interested in tuple spaces (typified by Linda, of which many implementations exist for modern programming languages) and other sorts of blackboard systems, which I think provide a very useful metaphor for characterizing scale-invariant concurrent distributed computation.
  • Reactive patterns of all sorts have grown in popularity and caché of late, such as FRP. Many of these reactive approaches are quite at home in a distributed context, where you might characterize computational services in general as operations that are triggered in response to data arriving in a local CRDT that matches a pattern or satisfies a query over the data being replicated in from other actors, similar to how triggers in relational databases are fired when a new result is found for a SQL query.

The bottom line is that it’s not at all clear yet which programming models are best suited for use with CRDTs…though I have to say, I do like my options.

With constraints come costs

Recalling the comparison I made earlier to programming languages enforcing progressively more constraints to minimize error and maximize ease of understanding, it’s worth asking what constraints CRDTs, semilattices, and other related approaches necessitate in order to yield their benefits. There are two, as far as I can tell:

It is imperative that CRDT implementations ensure that their operations adhere to the algebraic axioms of semilattices. (I think it is no accident that many CRDT implementors also happen to be very interested in things like property-based testing [e.g. QuickCheck, QuickCheck, test.check, double-check, and so on] and theorem provers like Coq, both of which are good tools for defining and verifying strict invariants like these.) This is definitional, but further means that operations and data types that are not naturally associative, commutative, and idempotent need to be recharacterized in terms of primitives that are. There are costs here (as with any abstraction, this boils down to a certain amount of computational, storage, and transmission overhead), but I’ll submit that the benefits are well worth it.

Secondly, if you want to benefit from the advantages offered by e.g. CRDTs, then you must use them to mediate all interactions between all actors in your distributed system. This implies a certain kind of boil-the-ocean sort of posture that can appear to be very costly, especially if you have a large investment in existing technologies that don’t yield similar benefits.

Given my background with and habituation to functional programming and immutable data structures, I think of this constraint as being similar: once convinced of their utility, you generally try to maximize the footprint of the “sane” world that uses pure functions operating over immutable data structures, and minimize the surface area of its interaction with the often quite incomprehensible world of effects and in-place mutation.

Likewise, the parts of a system that connect with and share state using CRDTs will be much easier to reason about and debug, so it is natural to want to maximize their use relative to the parts of your system that are tainted by things like usage of imperative APIs and less capable data representations. Both cases have a dynamic that favours those building new systems as opposed to those that maintain or steward large existing systems.

Totally aside from the actual constraints imposed by these approaches, a final word of caution might be that all of this is incredibly new. Even though they offer formalisations of mechanisms that have been employed in the past (very rarely, and never in principled, general-purpose ways), both the CALM theorem and CRDTs have only existed as discrete named concepts for the last few years. The first practical implementations of them are even newer than that. Significant bits of theory and implementation details are still up for grabs (e.g. CRDT “garbage collection” is a bit of a hot topic right now as these things go), so act accordingly. This is powerful stuff, and its potential is immense, but there will be rough waters between now and when CALM implementations and CRDTs are thought of as indispensable and obvious choices.

What do we want?

Earlier in this piece, I laid out what we want when we build a distributed system: communication and computation. I believe that CRDTs (or something like them), with their sound formal basis, give us a way towards building systems that hew close to those essential capabilities, avoiding the incidental complexity of APIs and much of the rest of the incidental complexity that comes along with modern typical distributed programming practice.

I believe these two essential capabilities can be sustained by services (and people!) reactively manipulating a shared substrate of replicated data. CRDTs appear to be an ideal, practical implementation of this substrate that connects disparate actors using reliable replication mechansisms, and thus allow for arbitrary computational models and flexible application and network topologies.

Resources

In addition to the links-as-citations that I have included throughout this piece, I could append here a thorough set of references. However, in doing so, I would effectively duplicate much of Chris Meiklejohn’s great work in compiling his Readings in Distributed Systems. I encourage you to absorb as much as you can from the resources referenced there if you have interest in these topics.

A harder alternative path would be to read about the well-grounded approaches I mentioned earlier; look up unfamiliar terminology, follow their bibliographies, and shout out on Twitter when you get stuck (there’s a good community of researchers and practitioners sharing references and experiences there):

Finally, keep an eye on the Quilt Project, the manifestation of my work in this and other areas. CRDTs are an essential part of its foundation.

Footnotes

  1. This dynamic alone significantly affects the social equation and economy between providers and consumers of network APIs…something I’m eager to discuss at length at some point, but not here, now. back
  2. Anyone claiming otherwise is implying that the status quo is acceptable, where the only parties that can reasonably build large, robust distributed systems are large organizations that have the resources (generally, many millions of dollars yearly) to spend on the engineering talent and bodies that appear to be necessary given the technologies currently in widespread use. back

Pragmatic discourse of Lebanese: Ye3neh? Jæʕne? « je veux dire »,, “I mean”

“What do you mean?”

Abstract of Layal Kanaan thesis

Jæʕne, marqueur issu des dialectes libanais, syrien, palestinien et égyptien, a été rencontré dans le cadre d’un travail de thèse consacré à l’étude de la reformulation dans les discours de Libanais francophones.

Il s’agit de discours menés en français au cours desquels les locuteurs emploient le marqueur de leur langue maternelle.

C’est la fréquence élevée du marqueur dans les données et dans les échanges des Libanais en général, ainsi que la diversité de ses emplois, qui ont justifié son étude.

Jæʕne présente le fonctionnement d’un marqueur discursif. Il peut être traduit, selon ses emplois, par c’est-à-direje veux diretu veux direquoi ou même par euh.

L’article présente l’évolution de l’unité verbale d’origine, propositionnelle, vers l’unité conversationnelle qui relève de la catégorie des marqueurs discursifs, et propose une typologie distributionnelle et fonctionnelle du marqueur dans le corpus, en prenant en compte la dimension prosodique.

Jæʕne partage certaines caractéristiques avec les verbes parenthétiques, à savoir son origine verbale, son figement morphologique, ses propriétés distributionnelles.

En revanche, sa particularité réside dans son figement à la troisième personne alors qu’on rencontre plus souvent les formes je trouvetu sais ou I mean, à la première et la deuxième personne (Thompson et Mulac, 1991).

Jæʕne dans le discours des Libanais : pragmaticalisation, distribution et emplois d’un marqueur discursif déverbal

Layal Kanaan

1. Introduction

  • 1 Différentes notations du marqueur sont attestées : jaʕnija’ni dans les travaux scientif (…)

1Nous nous intéressons à un marqueur discursif d’origine verbale employé de manière très fréquente dans certains dialectes de l’arabe (famille Levantin du Nord) : il s’agit de jæʕne. La notation adoptée correspond à la transcription phonétique de la prononciation en libanais du marqueur1.

2Dans la littérature, les travaux qui y font référence (Al-Batal, 1994 ; Kammensjö, 2005) l’abordent rapidement dans le cadre de l’étude des connecteurs de l’arabe. La classification qui en est proposée (connecteur explicatif et discourse filler) est incomplète. Jæʕne a aussi été observé par Traverso (2000) dans le dialecte syrien et abordé d’un point de vue interactionnel. L’auteure lui reconnaît une valeur de modalisation, « hedge » au sens de Lakoff (1972).

3Notre intérêt pour jæʕne est né dans le cadre d’un travail de thèse consacré à l’étude des reformulations dans des échanges menés en français entre des Libanais francophones.

Les 143 occurrences du marqueur relevées dans un corpus de 24 000 mots et leur lien intrinsèque avec le phénomène étudié (formulation / reformulation) nous ont amenée à accorder à jæʕne une place importante dans notre étude. Cette entreprise est aussi motivée par l’absence d’études qui y soient entièrement dédiées.

4Après une brève présentation du corpus, des locuteurs et de la situation linguistique au Liban, cet article aborde l’évolution du marqueur, appréhendée en termes de pragmaticalisation, à travers une démarche inductive. Il s’agit de la comparaison (en synchronie) entre les emplois de la forme verbale (en arabe classique et en libanais) et ceux de la forme « discursive », figée, employée non seulement dans le discours des locuteurs libanais dans leur langue maternelle mais aussi dans la pratique du français, comme l’attestent les occurrences dans le corpus.

En l’absence de données diachroniques, on s’attache à la vérification, dans le cas de jæʕne, de principes caractéristiques des unités pragmaticalisées. L’article dresse enfin une typologie des emplois du marqueur dans le corpus.

5Dans la mesure où les emplois étudiés correspondent aux occurrences dans le corpus recueilli, à côté de quelques exemples en arabe classique et en libanais (dialecte), la majorité des exemples est en français et jæʕne, dont l’emploi est extrêmement automatisé chez les locuteurs, est employé comme tel (en libanais).

6D’ailleurs, ses emplois dans le corpus constituent des phénomènes de contacts de langues auxquels nous nous sommes intéressée (Kanaan, 2013).

2. Corpus, locuteurs et terrain

7Notre « rencontre » avec jæʕne, devenu objet d’étude, s’est faite dans un corpus dont les données ont été récoltées expressément pour une thèse sur la reformulation dans les discussions en français de locuteurs libanais.

2.1. Corpus et locuteurs

8Il s’agit d’un corpus d’échanges oraux que nous avons constitué nous-même : mise en contact des locuteurs, établissement du dispositif, enregistrements et transcriptions. Les conventions de transcription adoptées figurent à la fin de cet article.

9Les locuteurs sont au nombre de 8 (4 hommes et 4 femmes). Ils sont tous libanais et ont fréquenté des écoles et lycées francophones. Ils ont entre 18 et 27 ans, sont étudiants ou jeunes actifs. Leur niveau d’études va du baccalauréat au master 2.

10Il s’agit d’apprenants de niveau avancé qui ont été scolarisés dans des établissements francophones où le français est la langue de scolarisation : enseigné à raison de 5 à 7 heures par semaine et langue véhiculaire des matières scientifiques. L’anglais, dans ces établissements, a le statut de première langue étrangère.

11Avec une moyenne d’âge de 23 ans, les locuteurs ont « achevé » l’apprentissage scolaire du français mais sont encore en contact avec cette langue dans leurs études universitaires ainsi que dans leur vie quotidienne (radio, télévision, publicités, lectures).

12Pour chaque enregistrement (4 au total, d’environ 30 minutes chacun), nous avons réuni deux locuteurs et leur avons proposé de débattre de cette question, formulée par nous-même, approximativement de la manière suivante : « Que pensez-vous de votre situation en tant que jeunes Libanais par rapport aux études, au travail, au mariage, à votre liberté, à vos relations avec vos parents ? Pensez-vous émigrer ? ».

13Le sujet des interactions a fait l’objet d’une réflexion. Sachant que ces dernières allaient être provoquées pour l’enregistrement, il nous revenait de proposer un thème favorisant les échanges, un thème qui concerne de près les locuteurs.

  • 2 Corpus de langue parlée en interaction, université de Lyon II.

14Il s’agit, en somme, d’interactions en face à face sur un thème imposé. Notre dispositif s’identifie dans une certaine mesure à celui mis en place pour la constitution du corpus « Mode » dans CLAPI2 : réunion de deux locuteurs et formulation d’une consigne. Voir, par exemple, Kerbrat-Orecchioni (2005 : 105) :

Il s’agit d’une conversation […] « authentique » mais non totalement « naturelle » car sollicitée par l’analyste, sur la base d’une consigne ainsi formulée : « Nous faisons une enquête sur le vêtement et les jeunes. Nous recueillons des discussions sur ce thème. Alors pendant dix minutes vous allez vous entretenir librement avec votre vis-à-vis sur le vêtement ».

15Par ailleurs, nous avons adopté, pour la sélection des locuteurs, la technique de « l’ami d’un ami » résumée comme suit (Bourdieu, 1993 : 1395, cité par Gadet, 2003) :

[…] choisir les enquêtés parmi des gens de connaissance […]. La proximité sociale et la familiarité assurent en effet deux des conditions principales d’une communication « non violente ».

16Les enregistrements ont eu lieu dans différents endroits. Les trois premiers (Ji-Pe, Ke-Kl et Ta-Le) ont été effectués dans des appartements. Le quatrième (Fy-Ma) s’est déroulé dans un café.

2.2. Langues en présence au Liban

17À côté de l’arabe classique, langue officielle du pays, principalement écrite – et parlée dans les discours officiels –, langue de l’administration et de l’enseignement, trois langues principales cohabitent aujourd’hui au Liban : le dialecte libanais, le français et l’anglais.

  • 3 Un phénomène linguistique intéressant est apparu avec le développement des moyens de (…)

18L’arabe dialectal libanais est la langue de communication dans toutes les situations. Il est employé dans tous les usages informels quotidiens, que ce soit à la maison, dans la rue, les commerces, au travail ou à l’école, et même dans les médias audiovisuels. Il ne s’enseigne pas, et est rarement écrit3. Il présente aussi une diversification en plusieurs variétés régionales avec des différences aux niveaux phonologique et lexical, mais cela ne cause aucune difficulté dans l’intercompréhension. Il a le statut de langue maternelle.

19Dans le cadre de la vie quotidienne au Liban, il n’existe pas de « besoin » pour les langues étrangères. Autrement dit, en famille, avec les amis, chez les commerçants, la langue vernaculaire est le dialecte libanais. Les langues étrangères sont avant tout des langues véhiculaires nécessaires pour la scolarisation et la formation universitaire et, dans une moindre mesure, des langues vernaculaires pour les échanges (surtout écrits : documents professionnels, courriels) dans un cadre professionnel.

20Néanmoins, le dialecte libanais porte les traces des langues en présence dans le pays à travers l’emploi de mots comme : « bonjour », « bonsoir », « merci », « pardon », « toilettes », « hi », « thanks », « sorry », « bye » ; termes que même une personne peu ou non scolarisée intègre dans son discours – avec des adaptations phonologiques plus ou moins marquées ([r] roulé et dénasalisation).

21Cependant, des pratiques d’alternances codiques plus importantes se retrouvent dans le parler des jeunes, comme en témoignent ci-dessous certains exemples du corpus, malgré la consigne : « parler en français ».

  • 4 Comment vas-tu ?

22Un phénomène assez représentatif est la commercialisation de tee-shirts qui portent l’inscription : « Hi ! Kifak4 ? Ça va ? », une construction très usitée qui illustre les mélanges linguistiques auxquels se livrent plaisamment les Libanais.

3. De l’unité verbale à l’unité conversationnelle

3.1. Jaʕnibjæʕnejæʕne

3.1.1. Jaʕni : une forme verbale de l’arabe classique

23En arabe classique, jaʕni est une forme verbale. Il s’agit de la troisième personne, masculin singulier, de l’inaccompli du verbe ʕana (« signifier »).

24Jaʕni se traduit littéralement en français par « (il) signifie ».

[1] الدخول في زمن الصوم يعني الدخول في صراع روحي ضد الش
ʔad-duxûl fi zaman ʔaṣ-ṣawm jaʕni ʔad-dukhûl fi sirâʕ rûḥî ḍiḍ ʔaʃ-ʃar
L’entrée dans la période du carême signifie l’entrée dans un combat spirituel contre le mal.
(1re ligne extraite d’Internet : http://www.zenit.org/​article-2273?l=arabic)

25L’exemple [1] montre l’emploi à l’écrit, en arabe standard, de jaʕni, verbe conjugué précédé d’un groupe sujet et suivi d’un groupe complément. Jaʕnis’accorde avec le sujet comme dans :

[2] تربية الأطفال تعني أن تكون قدوة للاطفال قبل أن تكون صديقاً لهم
tarbijat ʔal-ʔaṭfâl taʕni ʔan takûna qudwatan li-l-ʔaṭfâl qabla ʔan takûna ṣadîqan lahum
L’éducation des enfants signifie que tu sois un modèle pour eux avant que tu ne sois leur ami.
(1re ligne extraite d’Internet : http://www.dw-world.de/​dw/​article/​0,,5078909,00.html)

Taʕni étant la forme verbale du féminin singulier.

26Jaʕni peut être précédé par un anaphorique « هذا » (haða) ou « ذلك » (ðalika) ou « ما » () (« ça », « cela », « ce qui ») et / ou suivi par « اݩ » (ʔan) (« que »), comme dans l’exemple précédent, pour donner « cela signifie », ou « signifie que », ou encore « cela signifie que », « ce qui signifie que ». L’anaphorique reprend ainsi un ou plusieurs énoncés ; jaʕnidemeure au masculin à la troisième personne du singulier. Cette construction permet à jaʕni d’élargir sa portée en reprenant une idée développée sur plusieurs énoncés.

[3] ما يثير الشكوك لديه هو معرفته بعلاقة ما بمحض الصدفة لم يكن على دراية بها. هذا يعني أنه من الضروري على كل فتاة أن تقيّم الشاب على أساس مدى وعيه وتفهمه
mâ jusir ʔaʃ-ʃukûk ladajh huwa maʕrifatuh bi-ʕalâqa mâ bi-maḥḍ ʔaṣ-ṣiḍfa lam jakun ʕala dirâja biha. Hâða jaʕni ʔannahu min ʔaḍ-ḍarûri ʕala kull fatât ʔan tuqajjim ʔaʃ-ʃâb ʕala ʔasâs madâ waʕjih wa tafahhumih.
Ce qui suscite les doutes chez lui [l’homme], c’est le fait qu’il découvre par hasard une relation qu’il ignorait. Cela signifie qu’il est essentiel que chaque fille puisse évaluer l’ouverture d’esprit et la compréhension de l’homme.

27La construction haða jaʕni ʔan(na) (« cela signifie que ») est considérée comme une « connective clause » (Kammensjö, 2005 : 127) permettant d’introduire une explication.

28Nous concluons ce point en précisant que jaʕni verbal est la seule forme attestée en arabe classique.

3.1.2. Bjæʕne et jæʕne en libanais : comparaison des formes et des emplois

29Dans le dialecte libanais, bjæʕne est la forme qui correspond à jaʕni. Le [b] s’affixe à toutes les formes verbales à la troisième personne de l’inaccompli dans leur passage dans ce dialecte (jaktub > bjæktob : « il écrit » ; jaʃrab > bjæʃrab : « il boit »), et le [a] se transforme en [æ]. Les terminaisons en [i] deviennent d’une manière générale [e].

30Bjæʕne est donc le masculin de la troisième personne du singulier de l’inaccompli, btæʕne étant la forme que prend le féminin. Les deux genres sont illustrés en [4] et [5]. Les exemples relèvent de l’oral :

[4] شو بيعني سكوته؟
ʃu bjæʕne skûto ?
Que signifie son silence ?
  • 5 Autrement dit, « rêver d’une mariée » ou « voir une mariée dans son rêve ». (…)
[5] عروس بالحلم بتعني ورتي
ʕarûs bəl-ḥələm btæʕne wərte
Une mariée dans un rêve5 signifie un héritage.
  • 6 Ainsi que dans d’autres dialectes de l’arabe (notamment en syrien), voi (…)

31À côté de ces emplois verbaux, nous retrouvons dans le dialecte libanais6, une forme qui s’apparente morphologiquement et sémantiquement à bjæʕne mais dont la nature est fondamentalement différente et les usages multiples, c’est jæʕne.

  • 7 C’est l’animateur qui s’adresse à deux de ses invités.
[6] مضطرين لأسباب لوجستية نبعدكن عن بعضكن يعني مأعدينكن قبال بعض
məḍtarrîn li asbêb logistijje nbaʕʕedkon ʕan baʕəḍkon jæʕne mʔaʕdînkon ʔbêl baʕed
Nous sommes obligés pour des raisons logistiques de vous éloigner l’un de l’autre jæʕne on vous a installés l’un en face de l’autre.
(1re ligne extraite d’une émission télévisée type talk-show7)
  • 8 Il s’agit d’un exemple pris « comme tel » sur Facebook où, tout comme pour les mails, (…)
[7] eh ne7na 3enna majless nouwweb w 3enna majles wouzara bass heik 7elwin yen7atto decor, lazizin ya3neh
Oui nous nous avons une Assemblée nationale et un gouvernement qui font un beau décor, ils sont mignons jæʕne.
(1re ligne extraite d’Internet : Facebook8)

32Cette même forme est celle que les locuteurs du corpus ont employée dans les échanges menés en français :

[8] Pe : maintenant je ne suis pas frustrEE + *jæʕne* euh certAINS désirs + ok sONT réalisés mais pAS tellement OK? mais je ne sens pas la frustration ++ ok?
  • 9 Non c’est que je m’excuse […].
[9] Pe : quand j’arrive à cette compa:gnie et ensuite ça va être mon rôle de prendre des employés peut-être que je ne vais pas appliquer cette méthode je- j’appliquerai la méthode-
Ji : tu as dit peut-être *jæʕne*
Pe : *laʔ* *ʔənno*9 je m’excuse c’est pas peut-être (rires)
[10] Ca : à trente ans tu vis ta sexualité *jæʕne* + au maximum ++

33La comparaison de bjæʕne dans [4] et [5] avec jæʕne dans [6] à [10] permet de relever :

  1. du point de vue phonologique, l’absence du [b] ;
  2. du point de vue morphologique, le figement à la troisième personne (perte des désinences verbales) ;
  3. du point de vue syntaxique :
    1. le changement de la construction < sujet – bjæʕne – complément > en un jæʕne employé entre deux énoncés en [6] et en [8], à la fin d’un énoncé en [7] et en [9], ou entre les constituants d’un même énoncé en [10] ;
    2. bjæʕne dans [4] et [5] est un constituant à part entière de la proposition alors que jæʕne dans les exemples [6] à [10] peut être supprimé sans que cela n’affecte la grammaticalité de la proposition.

34Les deux dernières observations (2 et 3 : 1 et 2) montrent que jæʕnerelève d’une autre catégorie que bjæʕne. Son comportement dans les exemples [6] à [10] répond aux caractéristiques de la catégorie générale des marqueurs discursifs (désormais MD).

35Dans la mesure où il n’existe pas de définition unique qui puisse rendre compte de tous les aspects caractéristiques des MD, nous proposons une liste de propriétés qui permettent de circonscrire au mieux les traits définitoires de la catégorie.

36Ainsi, les MD :

  1. ont une fonction connective : ils permettent de lier une unité textuelle au discours précédent (Fraser, 1999) ;
  2. ne sont pas intégrés syntaxiquement à l’énoncé, ils sont doncparenthétiques (Brinton, 2008 : 1) et ne contribuent pas au contenu propositionnel ;
  3. sont optionnels : leur suppression ne rend le texte ni agrammatical ni incompréhensible (Brinton, 1996 : 247) ;
  4. ont une fonction principalement pragmatique : ils fonctionnent comme des signaux guidant l’interprétation de l’interlocuteur (Aijmer et Simon-Vandenbergen, 2009), leur compréhension dépend uniquement de paramètres contextuels et cotextuels ;
  5. sont caractéristiques de l’oral (Dostie et Pusch, 2007) puisque la nature spontanée, informelle et fragmentée de l’échange nécessite une interprétation presque instantanée du discours (Mosegaard-Hansen, 1996 : 145) ; ils sont, de ce fait, très fréquents à l’oral ;
  6. jouent un rôle dans la cohérence du discours : ils permettent au locuteur d’organiser son discours et à l’interlocuteur de comprendre des segments distincts comme un tout (Erman et Kotsinas, 1993 : 81) ;
  7. ont une portée variable, c’est-à-dire que l’unité qui contient le marqueur peut revêtir plusieurs formes et que le marqueur peut s’appliquer à un discours plus ou moins long (Mosegaard-Hansen, 1996 : 106) ;
  8. fournissent, d’un point de vue sémantico-pragmatique, des instructions à l’interlocuteur sur la manière d’intégrer leur unité hôte dans une représentation mentale du discours, ils remplissent une fonction métadiscursive (Mosegaard-Hansen, 1998 : 236 ; Traugott, 2004 : 306).

3.2. La pragmaticalisation de jæʕne

37Nous partons de l’hypothèse selon laquelle le MD jæʕne est issu d’un phénomène de pragmaticalisation, définie comme l’évolution d’unités lexicales ou grammaticales donnant lieu à des unités conversationnelles « qui ne jouent [plus] un rôle sur le plan référentiel » (Dostie, 2004 : 27), autrement dit, des unités qui « émerge[nt] de la structure phrastique pour s’approprier un rôle au plan textuel et interpersonnel » (Dostie, 2006 : 2), qui marquent une prise de position métadiscursive par le locuteur.

  • 10 Dostie (2004 : 34), en se référant à Hopper (1991), précise qu’en l’absence (…)
  • 11 Il s’agit plus précisément de la décatégorisation, la paradigmatisation, la(…)

38En l’absence de données diachroniques, et en nous fondant sur une démarche défendue par Hopper (1991 : 20) et par Dostie (2004 : 34-45)10, l’étude du marqueur prend appui sur les principes11 qui guident la pragmaticalisation.

39Ces principes sont appréhendés ponctuellement dans le cadre de l’analyse et nous y reviendrons dans la conclusion pour en dresser un bilan.

40Nous nous penchons aussi sur les propriétés sémantiques des formes verbales d’origine qui peuvent constituer des prédispositions / motivations au / pour le changement.

3.2.1. Hypothèses sur l’origine et l’évolution de jæʕne

41À partir de la comparaison des formes et des emplois des unités verbales et déverbales (3.1 et 3.2, ci-dessus), nous avançons deux hypothèses quant à la forme d’origine de jæʕne, qui reposent sur une même conception de la pragmaticalisation du marqueur, i. e. évolution déterminée par la fixation des stratégies discursives dans des structures morphosyntaxiques, évolution d’éléments lexicaux employés dans des cotextes spécifiques et hautement contraints (Traugott et Heine, 1991 ; Hopper et Traugott, 1993 ; Brinton, 1996) :

Hypothèse 1 : jæʕne est un classicisme ; c’est jaʕni appartenant à l’arabe standard qui développe des emplois déverbaux à partir de son emploi dans une construction verbale connective : (haða) jaʕni (ʔanna) [« (cela) signifie (que) »].

Hypothèse 2 : jæʕne vient de bjæʕne du dialecte libanais et a perdu le [b] – par attrition phonologique – et ses désinences verbales à partir de son emploi dans la construction : (hajda) bjæʕne (ʔənno) [« (cela) signifie (que) »].

  • 12 Brinton (2008 : chap. 5) remet en cause cette hypothèse dans l’étude de (…)

42Ces deux hypothèses correspondent à une tendance de la recherche actuelle qui considère que les MD d’origine verbale, les verbes parenthétiques comme tu voistu saisI meanyou know, etc., trouvent leur origine dans des matrices propositionnelles à complément telles que « I mean (that) »12, « tu vois (que) » ou « (comme) tu vois » (Bolly, 2010).

43Il s’agit pour ces marqueurs, tout comme pour jæʕne, d’emplois déverbaux qui coexistent avec les formes verbales « d’origine ».

44Mais la pragmaticalisation ne dépend pas uniquement de paramètres « externes », i. e. syntaxiques et contextuels (hypothèses 1 et 2), elle est aussi liée aux propriétés « internes » de l’unité d’origine, c’est-à-dire à son sémantisme.

3.2.2. Sur le sémantisme de jaʕni / bjæʕne

45Des différentes études sur les MD d’origine verbale, il ressort que leur équivalent non discursif, à savoir la forme verbale d’« origine », appartient à la catégorie générale des verbes de cognition qui inclut les verbes de connaissance (« savoir », « comprendre », etc.), de perception (« entendre », « écouter », « voir », etc.), de parole (« dire », « parler », etc.), catégorie particulièrement sujette à la pragmaticalisation. D’après Dostie (2006 : 3) :

[…] il semble exister une forme d’affinité entre sens qui rendrait certains développements sinon prévisibles, du moins naturels. C’est comme si, en quelque sorte, certaines unités lexicales étaient prédisposées à se pragmaticaliser.

46Que l’unité d’origine soit jaʕni ou bjæʕne, on est en présence de la même unité réalisée phonologiquement de manière différente. Le sémantisme est donc le même : « signifier ».

47Dans < X bjæʕne Y > / < X jaʕni Y > (X signifie Y), le sémantisme de Y est posé comme équivalent à celui de X, ou explicite celui de X. Il s’agit surtout d’une équivalence mise en place et reconnue par le locuteur. Cette équivalence est prise au sens large car on est confronté à l’une des possibilités suivantes :

  • X et Y sont équivalents ;
  • Y est la définition de X ;
  • Y explique, élabore X ;
  • Y exemplifie X ;
  • Y récapitule, résume X ;
  • Y nomme X.

48Dans tous les cas, l’opération d’identification – au sens large – en discours est une opération cognitive. Selon Fuchs (1994 : 174) :

[…] l’établissement d’une relation de paraphrase se joue sur un autre terrain que celui de la langue : il s’agit d’une stratégie cognitivo-langagière des sujets qui procèdent à une identification momentanée des significations de chacun des deux énoncés, annulant les différences au profit des seules ressemblances.

49Bien qu’il s’agisse dans cette citation d’une remarque sur la paraphrase, cette vision peut être élargie et vérifiée quant aux divers emplois dejaʕni / bjæʕne qui appellent une opération cognitivo-langagière.

50Jaʕni et bjæʕne ont pour interprétation sémantique « signifier ». Ce sémantisme, « cognitif », leur permet de développer des emplois textuels et interpersonnels, donc de se pragmaticaliser.

  • 13 Par « explication de l’intention », on entend ici l’explication d’un acte par la form (…)
  • 14 « NN » pour « nonnatural sense » (Grice, 1957 : 378).

51Mais jaʕni (tout comme bjæʕne) a un autre emploi. Il introduit l’explication de l’intention13, emploi apparenté au concept de « meaningNN »14 de Grice (1957), qui distingue un sens qu’il appelle « naturel » et un autre « non naturel ». Le sens « non naturel » est défini comme suit (Grice, 1957 : 385) :

“A meantNN something by x” is (roughly) equivalent to “A intended the utterance of x to produce some effect in an audience by means of the recognition of this intention” ; and […] to ask what A meant is to ask for a specification of the intended effect.

52Considérons l’énoncé suivant :

[11] ﺇﻦ ﺮﺍﻴﺘﻤﻮﻨﻲ ﺼﺎﻤﻂﺎ فاعلموا ﺃﻦ ﺼﻤﻂﻲ ﻻ ﻴﻌﻨﻲ ﺠﻬﻟﻲ ﺒﻤﺍ ﻴﺪﻮﺮ ﺤﻮﻠﻲ ﻮﻠﻜﻦ ﻤﺍ ﻴﺪﻮﺮ ﺤﻮﻠﻲ ﻻ ﻴﺴﺗﺣﻕ الكلام
ʔin raʔaytumûni ṣâmitan fa-ʕlamou ʔanna ṣamti lâ jaʕni ʒahli bimâ jadûru ḥawli wa lâkin mâ yadûru ḥawli lâ jastaḥiq al-kalâm
Si vous m’avez vu silencieux, sachez que mon silence ne signifie pasmon ignorance de ce qui se passe autour de moi mais ce qui se passe autour de moi ne mérite pas la parole.
(1re ligne extraite d’Internet : http://ejabat.google.com/​ejabat/​thread?tid=79ece30735fbb7be)

Jaʕni est employé ici dans le sens de meanNN. En effet, il introduit l’explication d’une intention, celle que le scripteur a voulu exprimer (et dont il entend qu’elle soit comprise comme telle) par « son silence ». Cet énoncé peut être reformulé par : Ce qui est signifié par mon silence est : « ce qui se passe autour de moi ne mérite pas la parole » (voir Grice, 1957 : 378).

53Ainsi, deux sens principaux sont attribués aux emplois de jaʕni / bjæʕne : celui qui introduit l’explication du sens d’un mot, d’un énoncé ou d’une idée, et celui qui introduit, comme on vient de le voir, l’explication d’une intention.

3.2.3. De quelques motivations de l’évolution

  • 15 Le besoin d’expressivité renvoie à la volonté du locuteur d’agir sur son interlocuteu (…)

54La principale motivation du changement linguistique est le besoin d’expressivité15 des locuteurs. Meillet (1982 [1921] : 139) précise que « ce qui en provoque le début, c’est le besoin de parler avec force, le désir d’être expressif ». Traugott (1982), quant à elle, présente cette motivation par la finalité :

Propositionnel > textuel > expressif.

55Le besoin d’expressivité paraît encore plus évident dans la pragmaticalisation. Les unités qui se spécialisent dans cette fonction, i. e.les MD, et qui interviennent principalement aux niveaux de la cohésion discursive, du guidage de l’interprétation et de l’intersubjectivité, constituent des réponses conversationnelles efficaces aux besoins des sujets parlants.

56À la question : « pourquoi certaines unités se pragmaticalisent-elles et pas d’autres ? », la réponse demande de faire intervenir différents paramètres. Nous en avons déjà présenté deux, concernant l’unité d’origine :

  • son emploi dans des cotextes contraints,
  • son sémantisme cognitif.

S’y ajoute le vide (ou besoin) linguistique, en d’autres termes, ceux de Meillet (1982 [1921] : 133) : « la “grammaticalisation” de certains mots crée des formes neuves, introduit des catégories qui n’avaient pas d’expression linguistique ».

57Tel est le cas pour jæʕne qui constitue le seul élément de son paradigme dans le dialecte libanais. Il existe en arabe classique une particule explicative : [ʔaj], mais cette forme n’a pas été conservée dans les dialectes. Son emploi se différencie de celui de jæʕne en ce qu’elle relie des unités de même niveau : [ʔaj] explique un mot par un autre, un syntagme par un autre, un énoncé par un autre. Ce n’est pas le cas de jæʕne qui relie des unités de niveaux différents faisant preuve, de ce fait, d’un usage plus large.

58Ainsi, parmi les motivations à la base de l’émergence du MD jæʕne figure l’absence, dans le vocabulaire libanais, d’autres unités permettant le marquage du lien d’équivalence ou de reformulation.

3.2.4. Complexification sémantique et attrition phonologique de jaʕni / bjæʕne

  • 16 Nous renvoyons à la distinction opérée par Pop (2000) entre espace métadiscursif et e (…)

59Si le principe d’affaiblissement sémantique rend compte de manière efficace de trajectoires de grammaticalisation, il se révèle inadapté à la pragmaticalisation dans ses premiers stades. C’est pour cette raison que nous rejoignons Traugott (2004) et Dostie (2004) pour parler plutôt de complexification sémantique à travers le passage d’un sens lexical, référentiel à un sens pragmatique, abstrait, méta- voire paradiscursif (Pop, 2000)16.

60On considère les différents emplois de jæʕne (voir infra, 4.3.1 à 4.3.6) comme des actualisations différentes d’une même unité : il s’agit d’un cas de polysémie. Il existe pour tous ces emplois, même pour ceux qui présentent le plus d’opacité par rapport au sens lexical d’origine, un sens commun, sous forme d’instructions, qui s’actualise différemment selon les contextes.

61En ce qui concerne l’attrition phonologique du marqueur, elle est observée à deux niveaux. Le premier n’est valable que si l’hypothèse 2 (jæʕne vient de bjæʕne du libanais) est vérifiée. Si c’est le cas, il s’agirait de la perte du [b] de bjæʕne.

  • 17 Fricative, sonore, non emphatique.

62Le deuxième niveau d’attrition, et ce, quelle que soit la forme d’origine, est le relâchement, dans la prononciation du marqueur, de l’articulation transcrite [jæʕne] jusqu’à l’amuïssement articulatoire total de la pharyngale17 et de la semi-consonne prépalatale [j], i. e. [əne], en passant par [jəne].

63Dans les réalisations les plus relâchées, les occurrences du marqueur tendent à être imperceptibles. En effet, à chaque révision des transcriptions, nous relevions des omissions de notation de certaines occurrences de jæʕne ; ces dernières se fondent prosodiquement dans le flux du discours et seule une écoute très attentive et informée permet de les repérer.

3.2.5. Le figement de jæʕne

  • 18 Expression empruntée à Dostie (2004).

64Ce qui rend jæʕne particulier par rapport aux autres « MD déverbaux »18, c’est que, contrairement à la majorité de ces marqueurs figés à la première ou à la deuxième personne, il porte la marque de la troisième personne.

  • 19 Voir Dostie (2004 : 67-70) pour une présentation plus complète des MD d(…)

65L’étude de Thompson et Mulac (1991) montre qu’en anglais les MD issus de constructions verbales sont à 95 % à la première personne, à 4 % à la deuxième et à 1 % seulement à la troisième personne. En français aussi, la majorité des MD déverbaux s’est figée à la première ou à la deuxième personne : disonsje comprendsdis donctu sais, tu vois19, par exemple.

66Les marques de la première et de la deuxième personne constituent des facteurs qui favorisent la pragmaticalisation, la subjectivité et l’intersubjectivité étant des besoins communicationnels essentiels pour les locuteurs. L’unité pragmaticalisée véhicule une expression plus marquée de la subjectivité du locuteur (Dostie, 2004 ; Traugott, 1995 et 2002).

67Dans le cas de I mean, par exemple, l’accent est mis sur la modification du discours antérieur du locuteur lui-même, par opposition à you mean qui permet au locuteur de « modifier » le discours de son interlocuteur (Schiffrin, 1987 : 299). La particularité de jæʕne réside dans le fait que ses emplois recouvrent ces deux possibilités : la même forme est employée par le locuteur pour modifier son propre discours et celui de son interlocuteur. Autrement dit, jæʕne, en discours, acquiert une dimension interpersonnelle en dépit de l’absence du marquage de la première ou deuxième personne d’un point de vue morphologique.

4. Typologie des emplois de jæʕne

68L’origine lexicale de jæʕne (jaʕni ou bjæʕne : « il signifie ») explique un emploi principal de ce dernier comme marqueur de reformulation, emploi qui selon les cas sera rendu par « c’est-à-dire » ou « autrement dit ». À côté de cette fonction figure un large éventail d’emplois renvoyant à des degrés de pragmaticalisation plus ou moins avancés, dans lesquels jæʕne marque la correction, la continuation, le travail de formulation (hésitation, recherche lexicale, planification discursive), etc.

4.1. Méthodologie

69L’objectif est de définir les emplois de jæʕne d’un point de vue sémantico-pragmatique en prenant en compte les paramètres distributionnels et contextuels du marqueur.

70L’intérêt pour jæʕne a été suscité par ses nombreuses occurrences dans les débats du corpus recueilli, en particulier par celles qui marquent les opérations de reformulation. La présente étude est fondée sur des exemples extraits de nos enregistrements.

71En suivant une démarche inductive, nous avons observé les différents emplois de jæʕne en prenant en compte des paramètres prosodiques, distributionnels et sémantico-pragmatiques.

4.1.1. Paramètres prosodiques dans la discrimination des emplois de jæʕne

  • 20 Pour une étude approfondie de la prosodie de connecteurs discursifs et de son rôle da (…)
  • 21 Perception à l’oreille, fidèle à la perception spontanée des locuteurs (voir Dostie, (…)

72Bien qu’une analyse prosodique ait pleinement sa place à côté de l’analyse distributionnelle et pragmatique dans la discrimination des emplois du marqueur20, elle a été restreinte à une analyse perceptuelle21 de l’intonation dans la réalisation du marqueur afin de distinguer deux types :

  • continuative : une intonation légèrement montante annonçant une suite ou « remontée de F0 [fréquence fondamentale] en fin de segment […] manifestant que l’on n’a pas fini de s’exprimer » (Morel et Danon-Boileau, 1998 : 16) ;
  • conclusive : une intonation descendante avec effet de clôture ou « la chute conjointe et rapide (sans allongement) de l’intensité et de F0 à un niveau bas » (Morel et Danon-Boileau, 1998 : 16).

73Une attention particulière a été accordée à l’intonation de l’unité précédant et / ou suivant l’occurrence de jæʕne, principalement pour le cas où l’unité est inachevée.

74Les paramètres suivants se sont avérés indispensables dans la détermination des fonctions du marqueur :

  • la réalisation de jæʕne ;
  • l’allongement vocalique de la dernière syllabe ou l’absence d’allongement ;
    • 22 Si les pauses remplies renvoient toujours au travail de formulation (…)

    les pauses22 ou leur absence avant et / ou après jæʕne, ainsi que les autres marques du travail de formulation (désormais TdF) avant et / ou après le marqueur : pauses remplies (« euh »), allongement vocalique.

75Si les aspects sémantico-pragmatiques du marqueur et de son cotexte ont participé à sa compréhension, les fonctions de jæʕne dans le discours n’ont pu être déterminées sans le passage par l’écoute des enregistrements du corpus. Voici un exemple présentant trois occurrences du marqueur qui montre comment les paramètres prosodiques participent à la reconnaissance du rôle de jæʕne.

[12] Ji : je ne sais pas c’est un cas particulier *jæʕne:* euh c’est peut-être que: + euh + *jæʕne* ta famille à toi/ + te laisse faire des choses mais + euh si on parle en générAL + *jæʕne* la vie au Liban il y a des familles qui sont un peu euh ++ traditionnelles qui: euh ne donnent pas de liberté euh + aux FILLes + euh précisément

76En [12], la première occurrence du marqueur est intégrée prosodiquement à son cotexte gauche (absence de pause) et annonce, par son allongement vocalique et son contour intonatif continuatif, une suite, une reformulation, en marquant l’insuffisance informationnelle de la séquence qui précède ; cette première occurrence est suivie d’une pause remplie qui signale le TdF en cours.

77Le deuxième jæʕne apparaît après une séquence de TdF constituée d’un allongement vocalique, d’une pause, d’une pause pleine (« euh »), puis d’une pause. Intégré prosodiquement au cotexte droit et précédé d’une séquence de TdF, jæʕne marque la continuation du discours.

78La troisième occurrence nous intéresse particulièrement. À la lecture et sans le recours aux paramètres sémantico-pragmatiques des unités qui l’entourent, elle serait interprétée comme un emploi reformulatif :

si on parle en générAL + *jæʕne* la vie au Liban

« [L]a vie au Liban » apparaissant comme la reformulation de « si on parle en général ».

79Or, à l’écoute, l’interprétation diffère. En effet, l’absence de pause après « la vie au Liban » et le contour intonatif continuatif de cette unité excluent la lecture reformulative. La formulation de Ji « la vie au Liban il y a des familles » est à entendre comme « au Liban il y a des familles ». Les marques prosodiques permettent donc la compréhension de l’emploi dejæʕne comme continuateur et non comme marqueur de reformulation.

80La prosodie permet aussi la délimitation de l’unité hôte du marqueur et, de ce fait, la détermination de sa position par rapport à cette unité. L’intonation continuative place jæʕne au début de l’unité qui le suit alors que son intonation conclusive marque sa position finale.

81Par ailleurs, pour certaines occurrences, on peut reconnaître la fonction du marqueur alors que sa portée est absente :

[13] Pe : […] #1 [xxx] tu peux commencer- #
Ji : #2 non mais- mais déjà tu as vingt-cinq ans tu as vingt-six vingt-sept*jæʕne*– #
Pe : tu peux commencer par payer le premier versement par exemple et chaque fois tu payes de plus/
[14] Ta : #2 oui le problème oui que- # que il y a parfois des gens qui se marient même dans- *jæʕne* dans la jeune génération maintenant/ + ils se marient car son père veut qu’elle se marie de cette personne parce qu’il a- parce qu’il est riche et tout ça et- *jæʕne*-
Le : il y aura [teʒur] d- il y aura toujours ces exceptions […]
  • 23 Les occurrences observées ici sont marquées en gras.

82En effet, les occurrences de jæʕne en [13] et en [14]23 marquent proactivement un segment de discours dont la formulation a été interrompue par la prise de parole de l’interlocuteur, mais à travers l’écoute, on est en mesure d’y reconnaître une intention de reformulation. Une étude instrumentale de la prosodie du marqueur présenterait un intérêt majeur pour la discrimination de ses emplois.

4.2. La distribution de jæʕne

4.2.1. Méthodologie

La distribution d’un élément sera définie comme la somme de tous les environnements de cet élément. (Harris, 1970 : 14)

83La notion de distribution pour un MD diffère sensiblement de celle applicable aux catégories lexicales. Les MD sont des unités qui ne sont pas grammaticalement dépendantes du verbe constructeur. Ils ne participent pas au contenu propositionnel de leur unité hôte.

84La distribution d’un MD doit être examinée à partir de la portée, scope, du marqueur lui-même. En effet, pour une description pertinente des « environnements » du marqueur, il importe de déterminer le segment du discours sur lequel il porte, d’une part, et d’observer sa position par rapport à cette unité, d’autre part.

85De plus, la prise en compte du segment précédant le marqueur (S1 dans : < S1 – marqueur – S2 >), si celui-ci est en position initiale, ou de celui qui précède son unité hôte lorsque ce dernier est en position finale (S1 dans : < S1 – S2 – marqueur >), constitue une information importante. En effet, Schiffrin, qui définit les marqueurs discursifs comme « sequentially dependent elements which bracket units of talk » (1987 : 31), précise (1987 : 37) :

[…] it is important to note that brackets look simultaneously forward and backward – that the beginning of one unit is the end of another and vice versa.

  • 24 Par doublement cataphorique, nous caractérisons le cas où le marqueur est employé à l (…)

86De ce fait, qu’il soit proactif ou rétroactif, le marqueur présente un double caractère : respectivement, anaphorique / cataphorique et doublement cataphorique24.

87La prise en compte de la séquence antérieure repose sur des paramètres sémantiques et pragmatiques. Elle ne peut pas être définie à partir d’indices formels propres mais nécessite l’interprétation du récepteur. Le cotexte de gauche, précédant l’unité hôte, permet d’affiner l’observation distributionnelle du marqueur.

88On a parlé de séquences, de cotexte de gauche et d’unité hôte, or ces délimitations ne vont pas de soi et supposent, à côté de la dimension pragmatique, une conception préalable des unités constitutives d’un texteoral. Selon Le Goffic (2008 : 329) :

La recherche des unités d’un texte est à la fois une nécessité évidente […] et une source de contradictions inévitables : qui dit unité dit individualisation, indépendance, autonomie, mais qui dit texte (tissu) dit interdépendance des parties.

89Cette question appelle une conception de la syntaxe de l’oral et la notion de macrosyntaxe étudiée par le Groupe aixois de recherche en syntaxe.

90La macrosyntaxe renvoie aux « relations qu’on ne peut décrire à partir des rections de catégories grammaticales » (Blanche-Benveniste (éd.), 1991 : 113). Il s’agit d’« une organisation qui n’est pas fondée sur des catégories mais sur des unités d’ordre pragmatique, sémantique et prosodique » (Blanche-Benveniste, 2008 : 308) et dont les éléments constitutifs sont les énoncés.

91Dans l’étude distributionnelle de jæʕne, on considère comme unité de base l’énoncé, en séparant les occurrences du marqueur aux frontières des énoncés de celles qui apparaissent au sein de l’énoncé et qui portent sur une partie de ses constituants (partie d’énoncé, syntagme, mot). On parlera alors de deux niveaux : le niveau textuel et le niveau de l’énoncé.

4.2.2. Résultats de l’analyse distributionnelle de jæʕne

92Jæʕne au niveau textuel présente une distribution variée. Il a sous sa portée des segments du discours de longueur variable(phrase / paragraphe).

[15] Ji : […] ici au LibAN + il n’y a rien/ il n’y a pas de sécurité sociale si quelque chose t’arrive et tu dois aller à l’hôpital tu dois payer beaucoup d’argen::t t- les- euh toutes les choses il y a beaucoup de saletés sur les rue:s/ il y a-
Pe : *la2* mais quand même il y a::-
Ji : #1 *jæʕne* ce qui est bien c’est seulement la famille #
Pe : #2 il y a la sécurité sociale + #
Ji : c’est- c’est tout::

93Il peut occuper une position initiale, comme en [15], ou finale, en [16], par rapport à l’énoncé qui constitue sa portée.

[16] Ji : mais il est marié/ à quel an il est marié ?
Pe : à quel an il est ma-? non il est- il était marié ici au Liban et puis ils- euh #1 ils sont partis #
Ji : #2 ah il- il est allé ta:rd *jæʕne*

94Son apparition est indépendante de la modalité (déclarative, interrogative en [17], exclamative en [18]) de son énoncé hôte.

[17] Pe : quand je termine mes études
Ji : *jæʕne* à vingt-trois ans ? vingt-deux ?
[18] Ji : euh je sais que mon pays a besoin de moi ++ mais parfois aussi le pays i- *jæʕne* i- il doit m’offrIR des choses + on ne peut pas continuer comme ça/ *jæʕne* +

95Il peut, enfin, introduire des énoncés d’arrière-plan insérés entre le début d’un énoncé inachevé et sa répétition / son achèvement, jouant ainsi un rôle dans la structuration et dans l’organisation au niveau informationnel, comme dans [19].

  • 25 Afin de faciliter la compréhension de cet exemple, nous avons marqué par (…)
[19] Ta : […] car ces jeunes- ces jeunes gens/ ne savent pas comment exprimer c’est-à-dire leurs sentimen::ts « est-ce que » /- *jæʕne* après quand ils vont grandi:r « est-ce que j’aime cette fille? est-ce que je n’aime pas cette fille? »25 c’est-à-dire ils ne vont pas avoir euh **wiseness enough**

96La locutrice interrompt son énoncé en cours et introduit avec l’emploi dejæʕne l’information qu’elle juge nécessaire pour la compréhension de l’énoncé amorcé, avant de le reprendre.

97Au niveau de l’énoncé, la distribution du marqueur n’est pas moins diverse. Sa portée variable embrasse aussi bien le lexique que lessyntagmes. De manière moins fréquente qu’au niveau textuel, jæʕne peut, sur le plan de l’énoncé, être en position finale (voir l’exemple [20]) et cadrer de manière rétroactive des segments de ce dernier.

[20] Ma : euh une ami::e non + une copine/ ++ je n’en ai pas + mais il y a beaucoup d’amies qui:: comm- ++ comment dire/ + pour tous les jours*jæʕne*

98Pour les deux niveaux, l’observation de la distribution du marqueur n’a pu s’affranchir des éléments prosodiques qui ont servi, d’une part, à déterminer la position du marqueur (initale / finale par rapport à son unité hôte) et, d’autre part, en prenant en compte les pauses, à distinguer pour une même distribution du marqueur des fonctionnements différents.

99Jæʕne occupe trois places par rapport à l’énoncé. En position initiale, il fonctionne à un niveau textuel tandis qu’en position médiane, il fonctionne au niveau des éléments constitutifs de l’énoncé. Jæʕne en position finale peut relever de l’un ou l’autre de ces niveaux.

100Toujours pour les deux niveaux, on a relevé l’emploi de jæʕne dans des espaces de rupture trahissant des difficultés dans la formulation : énoncés inachevés et marques de TdF – i. e. pauses silencieuses et pauses remplies (« euh », allongements syllabiques) – constituent fréquemment le cotexte gauche du marqueur.

101En somme, jæʕne se caractérise par une grande liberté syntaxique. Les exemples analysés montrent la « flexibilité » du marqueur, notamment en ce qui concerne sa portée et sa position.

4.3. Les emplois de jæʕne

102L’analyse des emplois de jæʕne est délicate. En effet, ce marqueur présente une distribution variée, il a des emplois divers, pour certains difficiles à cerner ; entre continuation, reformulation, correction, approximation, les frontières s’estompent avec l’emploi de ce MD qui, tout en participant à l’avancement du discours, y indique continûment des retours.

4.3.1. Reformulation

103Le sens de jaʕni / bjæʕne révèle deux sémantismes principaux : l’explication du sens et l’explication d’intention. Jæʕne hérite parfaitement des sémantismes de sa (ses) forme(s) verbale(s) d’origine et permet de marquer des reformulations de nature différente dans le prolongement de ses deux potentialités.

[21] Fy : quand- quand on est- quand on est + très riche
Ma : mh mh
Fy : mh:: euh ++ on devient:: comme: euh + on- on- on aime l’argent beaucoup et on a beaucoup d’attachement à cet argent + quand on est/ + très euh
Ma : on devient avare *jæʕne* ?
[22] Ma : tu trouves qu’on:: s- qu’on ne peut pas s’adapter + ensemble
Fy : oui on peut pas:: ++
Ma : *jæʕne* l’un- l’un ne peut pas s’adapter avec le- l’autre […]

104Jæʕne marque les reformulations au niveau textuel et au niveau de l’énoncé. Au niveau textuel, il peut être en position initiale ou en position finale alors qu’au niveau de l’énoncé, il ne peut marquer la reformulation qu’en position finale. Il peut être suivi de pauses et de marques du TdF ou précédé par ces dernières. En position initiale, son regroupement prosodique avec l’un ou l’autre de ses cotextes marque, s’il est regroupé à gauche, la projection d’une reformulation, sinon il est précédé de pauses ou autres signes de TdF et il marque le segment de droite, avec lequel il est regroupé comme étant la reformulation du discours précédent.

4.3.2. Correction

105Bien que la correction fasse partie de la reformulation telle que nous la concevons dans le cadre de ce travail, nous avons séparé les emplois dejæʕne qui marquent la reformulation de ceux qui marquent la correction. Dans les deux cas, jæʕne indique un retour sur le « dit » et l’opération intervient sur le (segment de) discours antérieur. Mais si, dans la reformulation, le deuxième segment présente une élaboration du premier, dans la correction, il s’agit d’un remplacement.

[23] Ta : car il y a encore des parents même qui sont- *jæʕne* des couples qui sont même jeunes ils ont- j’ai un cousin à moi/ il est- *jæʕne* il est un- *jæʕne* il est un couple- *jæʕne* lui et sa femme ils sont jeunes
[23’] Ta : car il y a encore des parents même qui sont-
                *jæʕne* des couples qui sont même jeunes ils ont-
j’ai un cousin à moi/ il est-
                *jæʕne* il est un-
                *jæʕne* il est un couple-
                *jæʕne* lui et sa femme ils sont jeunes

106Jæʕne marque une série de corrections en [23]. La représentation en grille (Blanche-Benveniste (éd.), 1991) en [23’] met en évidence le TdF. La première correction porte sur le choix lexical de parents / couple et opère un remplacement. Les hésitations et corrections qui suivent sont les conséquences d’une même difficulté : le passage du générique au spécifique, qui présente une antinomie entre la valeur spécifique construite par « un cousin à moi » et la valeur générique du contexte précédent (voir Morel et Danon-Boileau, 1998 : 75-76).

107En [24], après une hésitation, le locuteur formule « habilités » et annonce immédiatement une rectification du terme anglais adapté phonétiquement.

[24] Ji : […] ici on te prend pas pour tes :: euh ha- habilités *jæʕne* pour tes **abilities** *jæʕne* pour tes : euh
Pe : capacités
Ji : capacité ::s éducatives mai :s pour tes- euh pour tes relations publiques et pour les gens que tu connaisses et c’est tout/
  • 26 Ce type de séquence correspond à ce que De Pietro, Matthey et Py (1989) appellent (…)

108Il y a reprise du cotexte droit « pour tes », structure d’accueil du terme recherché. La correction est proposée par l’interlocutrice – « capacités » –, reprise (ratifiée) et complétée par le locuteur avant qu’il ne poursuive son discours26.

4.3.3. Réparation d’arrière-plan

  • 27 Terme emprunté à l’ethnométhodologie (Schegloff, Jefferson et Sacks, 1977 ; Schegloff(…)

109Le marqueur peut apparaître au début de séquences discursives du type « background-repair » (Schiffrin, 1987 : 300). Cet emploi peut être rapproché des emplois d’édition dans la mesure où jæʕne apparaît après une rupture discursive qui correspond à un moment « cognitif » de prise de conscience d’un « problème » ou « trouble source »27. Il s’agit plus particulièrement dans ces cas, et selon les données dont nous disposons, d’un manque qui relève du niveau informationnel.

[25] Ta : IL Y A DES filles parfois qui pren:d l- *jæʕne* moi ce que je vois qu’il y a toujou- *jæʕne* qu’il y aura toujours des exceptions mais la plupart des filles il prend les extrêmes comment dire/ toujours/

110La réparation d’arrière-plan se présente sous la forme d’un énoncé ou d’une proposition qui rompt l’énoncé en cours et qui est suivi par la reprise / répétition de l’amorce en vue de sa complétion. Ce type d’opération relève de l’organisation informationnelle dans le discours dont la visée générale est l’intercompréhension. Pour revenir à nos exemples dans lesquels jæʕne signale ce type de réparation, nous observons qu’il marque, en position initiale, les segments d’« arrière-plan » et relevons les balises d’organisation suivantes :

  • l’emploi de jæʕne,
  • la répétition du segment amorcé.

Ces balises guident l’interlocuteur sur la manière de recevoir l’information ainsi mise en place, et dans sa compréhension.

111Jæʕne occupe toujours une position initiale dans ce type d’opérations. Il peut être suivi et / ou précédé de pauses brèves. Parmi les occurrences de cet emploi du MD dans le corpus, aucune n’est précédée ni suivie de marques de TdF et une seule est précédée d’une pause très brève.

4.3.4. Approximation et rétrocomplétude

112Nous avons dégagé deux principaux types d’emplois de jæʕne rétroactif (non reformulatif, non correctif), dont des exemples représentatifs figurent respectivement en [26] ainsi qu’en [27] et [28], la singularité en [27] et [28] tenant à ce que jæʕne apparaît à la fin d’énoncés interrogatifs.

  • 28 Certaines séquences, marquées par « […] », ont été supprimées afin d’assurer une meil (…)
[26] Ji : oui mai:s- mais pour- pour avoir- pour avoir la chANce d’entrer dans cette compagnie il faut avoir des relations publiques/ *jæʕne* tOUt ça c’est comme des choses un peu politiques/ *jæʕne* + tu ne peux pas entrer- entrer dans euh un dans un tra- tu ne peux pas avoir un travail ou bien être dans une compagnie euh très bien […]28 au Liban si tu n’as pas de- *jæʕne* ici on te prend pas pour tes […] capacité::s éducatives mai:s pour tes- euh pour tes relations publiques et pour les gens que tu connaisses et c’est tout/
  • 29 L’emploi de « mettre » dans cette séquence est à comprendre dans le sens d’« embauche(…)
[27] Ji : #2 je doute + je doute fort # moi- moi je vais mettre29 mon frère parce que ici- + euh comment tu n- tu ne dois pas mettre ton #1 frè:re? et qu’est-ce que va dire ta mère?#
Pe : #2 c’est-à-dire tu ne dois pas t’occuper des autres si tu fais # qu’est-ce que-
Ji : qu’est-ce que va dire ton père *jæʕne*? pourquoi tu n’as pas mis ton frère pourquoi [xxx] ?
[28] Ji : tu- tu peux vivre avec- euh avec un IL au Liban? non
Pe : #1 non je ne peux pas #
Ji : #2 avec un- tu peux pas # tu peux pas parce que- à cause de #1 ta famille de tes parents #
Pe : #2 à cause des coutumes des # traditions et tout ça mais- euh
Ji : toi tu as quelque chose contre ça *jæʕne*? euh si- s’il n’y a pas ces traditions tu as qu- quelque chose contre vivre- + contre le fait de vivre avec un homme […]

113Une difficulté majeure s’est posée au regard de la totalité des emplois illustrés ci-dessus, quant à la détermination de la portée du marqueur. Si nous avons la certitude, et ce, grâce à la prosodie, que le marqueur est rétroactif (intonation descendante), nous n’avons pas pu délimiter le segment du discours sur lequel il porte. Par exemple, en [26], [27] et [28], porte-t-il sur tout l’énoncé ? Ou seulement sur une partie de ce dernier ? Si la réponse à la deuxième question est positive, reste à délimiter les éléments de l’énoncé sur lesquels il porte et à déterminer comment il les modifie.

114Nous pensons que le marqueur en [26] porte sur un segment de l’énoncé, plus particulièrement sur le mot « politiques », et qu’il le marque comme étant l’élément qui complète l’énoncé, surtout que son emploi vient après une suite d’approximations : « c’est comme des choses un peu ».Jæʕne actualise également une valeur de « hedge » (Lakoff, 1972) : comme si le locuteur marquait le mot comme étant approximatif, sollicitant, par ce fait même, l’indulgence de l’interlocuteur et l’acceptation de son emploi. Tel est le cas aussi d’occurrences déjà observées dans l’étude distributionnelle du marqueur :

[29] Ma : euh une ami::e non + une copine/ ++ je n’en ai pas + mais il y a beaucoup d’amies qui:: comm- ++ comment dire/ + pour tous les jours*jæʕne* + *ʔənno* on sort tous les jours mais il n’y a pas de- UNE qui est spécifique
  • 30 Sur la notion deparaphrase discursive, voir Fuchs (1994).

115Il est intéressant de relever, pour la totalité des exemples présentés ci-dessus, les reformulations systématiques des séquences marquées rétroactivement par jæʕne. En effet, en [26], l’opération mise en place par le locuteur semble être déclenchée par l’emploi approximatif du terme « politiques », qui ne satisfait apparemment pas son intention de communication. En [27], il s’agit d’un discours direct attribué au « père de l’interlocutrice » et qui explicite la valeur de « qu’est-ce que va dire ton père ? ». Ce dernier segment se révèle non pas comme une question qui attend une réponse mais comme une mise en scène de l’attitude du « père », c’est-à-dire celle de reproche. En [28], c’est le déictique « ça » qui fait l’objet d’une paraphrase discursive30 et en [29], la reformulation permet une réorganisation de la formulation précédente, hésitante et déployée par à-coups.

116En position finale, jæʕne peut marquer l’aboutissement d’une recherche lexicale, mais il s’avère que la recherche en question n’aboutit pas toujours à un élément satisfaisant, d’où les reformulations observées en [26] et en [29]. Dans des emplois assez similaires, jæʕne marque aussi l’insatisfaction à l’issue d’une formulation (i. e. un énoncé ou des constituants de ce dernier), dans le sens où il invite à considérer qu’une reformulation est possible (exemple [27]) sans que celle-ci ait lieu explicitement.

4.3.5. Continuation

117Par jæʕne continuateur, on entend un emploi qui signale la poursuite par le locuteur de son propre discours.

[30] Ji : je ne sais pas c’est un cas particulier *jæʕne* euh c’est peut-être que:+ euh + *jæʕne* ta famille à toi/ + te laisse faire des choses
  • 31 Celle que nous avons marquée en gras.

118L’occurrence de jæʕne en [30]31 marque la poursuite de l’énoncé suspendu par un TdF. Comme le montre cet exemple, jæʕne continuateur est regroupé prosodiquement avec son cotexte droit et précédé par des marques de TdF (« euh », pauses et allongements vocaliques).

119La principale différence entre jæʕne continuateur et jæʕne marqueur de TdF se situe au niveau de la présence de pauses ou de marques de TdF :jæʕne continuateur n’en admet pas dans son cotexte droit immédiat.

4.3.6. Travail de formulation

120Dans les travaux d’Al-Batal (1994) et de Kammensjö (2005), jæʕne a été classé dans la catégorie des discourse fillers, dans le sens où il est employé dans les moments de « panne » et de « planification » discursives. Or, dans le cadre de notre étude, cette fonction de jæʕne est définie comme étant UNE parmi celles que peut avoir le marqueur. Dans ce rôle, jæʕne a été désigné comme « marqueur de TdF ». L’exemple [31] montre la cooccurrence du marqueur avec d’autres signes de ce type de travail.

[31] Ji : je trouve que même si on fait beaucoup de travail ici + *jæʕne* euh les conditions restent les mEmes #1 il y a- il y a beaucoup de gens qui pensent à EUX-mêmes #

121Dans cet exemple, le marqueur peut être identifié à une pause remplie, à la manière de « euh », cette dernière étant définie comme l’indice d’une « intention de continuation » (Morel et Danon-Boileau, 1998 : 82). Mais, même si jæʕne partage avec « euh » la fonction de marque de TdF, il ne se réduit pas pour autant à un son, à un discourse fillerJæʕne a une signification qu’il actualise dans le contexte particulier de son emploi dans le cadre du TdF. Il signale à l’interlocuteur l’effort en cours pour produire une forme correspondant à l’intention de communication du locuteur.

4.3.7. Bilan des emplois de jæʕne

  • 32 Ou une situation antérieure, bien que le corpus n’en fournisse aucun exemple.

122Il est apparu, comme dénominateur commun à tous ses emplois, quejæʕne suppose toujours un discours antérieur32 au segment du discours qu’il marque. Autrement dit, il assigne à son unité hôte un rôle ou une valeur à inférer du discours antérieur. Tout en modifiant / ajustant / faisant avancer le discours, il maintient l’attention du locuteur et par conséquent, celle de son interlocuteur, sur « ce qui a été dit ». Mais jæʕne n’est pas pour autant tourné vers l’interlocuteur. Son emploi montre une centration sur le discours (une fonction « méta »), même si l’objectif ultime de tout discours est l’échange, l’action sur autrui, l’intercompréhension.

123Jæʕne présente une unité thématique avec son cotexte gauche, dans le sens où il ne peut pas marquer un changement de « sujet », mais, au contraire, une progression thématique qui s’inscrit dans la même orientation argumentative que le discours antérieur.

124Dans tous ses emplois, jæʕne montre un fonctionnement étroitement lié à la formulation, que ce soit dans l’entreprise de son amélioration / réorganisation (reformulation), dans sa correction, dans ses difficultés (TdF) ainsi que dans ses aboutissements (continuation et approximation). Il est donc intimement lié à des problèmes de formulation, des formulations « qui ne vont pas de soi » (Authier-Revuz, 1995).

125Ce qui s’avère commun à tous les emplois de jæʕne et qui constitue une piste sérieuse dans une future entreprise visant à déterminer la signification du morphème, est l’idée d’insatisfaction qu’il actualise de manière variée selon ses emplois. Cette insatisfaction concerne des éléments de la situation antérieure. Pour la totalité des occurrences relevées dans le corpus, il s’agit de la situation discursive, mais pour des emplois attestés dans le dialecte libanais, il peut s’agir d’une situation extradiscursive, par exemple :

[32] A fait tomber un vase.
B : jæʕne ! + tu es le roi des bêtises !

126L’exemple [32] montre un emploi de jæʕne qui diffère sensiblement des emplois observés dans le corpus. Il en est de même pour jæʕne en [33] :

[33] A : comment se passe ton séjour ?
B : jæʕne

où il fonctionne comme un adverbe et peut être traduit par « couci-couça ».

127Dans chacun de ces deux exemples, jæʕne exprime l’attitude du locuteur, liée à l’insatisfaction.

Typologie des emplois de jæʕne dans le corpus

  • 33 Pauses remplies : « euh », allongements syllabiques.
Type Fonction Distribution Position Pauses et autres marques de TdF33
T 1 Marqueur de reformulation textuel initiale ou finale admet avant et / ou après sauf en position finale (jamais avant)
énoncé initiale
T 2 Correction textuel initiale admet avant et / ou après sauf en position finale (jamais avant)
énoncé initiale ou finale
T 3 Réparation d’arrière-plan textuel
énoncé
initiale jamais avant
T 4 Approximation
Rétrocomplétude
textuel
énoncé
finale jamais avant
T 5 Continuateur énoncé initiale toujours avant
jamais après
T 6 TdF énoncé initiale nécessite avant et / ou après

5. Conclusion

128Nous avons montré, à travers une démarche inductive comparant les formes jaʕnibjæʕnejæʕne et leurs fonctions, et en nous inscrivant dans le cadre théorique de la pragmaticalisation (Dostie, 2004 ; Erman et Kotsinas, 1993 ; Hopper et Traugott, 1993), que jæʕne est un marqueur discursif déverbal dont les emplois se sont développés à partir de l’emploi de l’unité verbale d’origine dans un cotexte contraint : une construction conjonctive. Nous avons relevé pour jæʕne une externalisation de la proposition et une distribution échelonnée.

  • 34 Ces principes sont présentés par Dostie (2004). Le principe de superposition renvoie au f (…)

129Notre approche de la pragmaticalisation du marqueur s’est principalement fondée sur l’observation de principes qui guident les changements linguistiques. En effet, à côté de la décatégorisation (perte des marques morphologiques et des particularités syntaxiques de la forme et de la catégorie verbales, et acquisition des propriétés de la catégorie des MD) et de la paradigmatisation (intégration de jæʕne à de nouveaux paradigmes, ceux des marqueurs de formulation / reformulation), nous avons observé les principes de superposition (layering) et de persistance34et à travers les différents emplois du marqueur – en synchronie et qui coexistent avec la forme verbale d’origine (bjæʕne / jaʕni) – parmi lesquels figure celui qui reflète le sens de l’unité d’origine, à savoir jæʕne marqueur de reformulation. Nous avons aussi relevé, d’un point de vue phonologique, une attrition dans certaines réalisations du marqueur.

130Il est intéressant de souligner qu’en dépit de son figement à la troisième personne, jæʕne partage des emplois avec la construction verbale parenthétique « je veux dire », témoignant ainsi de l’acquisition d’une dimension expressive, le besoin d’expressivité étant d’ailleurs à la base des phénomènes d’évolution.

131La variation synchronique des emplois de jæʕne mise en évidence à travers la typologie de ses différents emplois dans le corpus est la trace d’un processus diachronique de pragmaticalisation. On pourrait penser que chacun de ces emplois correspond à un degré différent de pragmaticalisation.

132Nous pensons aussi que ce processus est loin d’être achevé au vu du potentiel multifonctionnel dont fait preuve le marqueur. En ce sens, nous rejoignons Erman (2001 : 1357) qui précise :

[…] once this process has started the doors are open for meaning shift and language change to take place.

133Par ailleurs, on relève les limites d’une typologie qui se fonde sur un corpus qui n’est pas représentatif de tous les emplois du marqueur et oùjæʕne apparaît comme un phénomène d’alternance codique. Néanmoins, elle rend compte des emplois du marqueur dans le corpus et constitue un point de départ pour une typologie exhaustive qui requerrait des interactions en libanais.

134Enfin, en lien avec le contexte dans lequel les occurrences ont été observées, un grand nombre d’emplois de jæʕne, ceux qui marquent le TdF, sont typiques des situations exolingues où le marqueur traduit les difficultés quant à la formulation. On pourrait penser que ce type d’emploi est lié, de manière générale, à une situation cognitive difficile.

Conventions de transcription

+ pause courte (++ moyenne, +++ longue)
: allongement vocalique (::, :::, selon la durée)
/ intonation montante
? intonation interrogative
j- j’ai le tiret marque une troncature (un autre exemple : il faut les remp- remplacer ou il par- il s’en allait)
NON mot prononcé de façon accentuée
[xxx][stazje] mot ou segment inaudible, ou incompréhensible, ou transcription phonétique
#1 blabla##2 blibli# chevauchement des tours de parole
euh hésitation, pause remplie
(rires) commentaire du transcripteur pour caractériser des phénomènes non verbaux
*laʔ* mot ou segment en libanais (transcription phonétique)
**already** mot ou segment en anglais

Haut de page

Bibliographie

Only 12 words to define Entrepreneurship?

BILL MURPHY JR updated on April 24, 2014

There’s a definition of entrepreneurship that has changed how I think about the way people choose their paths in life. It helped me to build a thriving business and find all kinds of great new experiences. Heck, it even helped me to meet my wife.

I believe it can have the same kind of positive impact for you, if you’re willing to try to put it into practice:

Entrepreneurship is the pursuit of opportunity without regard to resources currently controlled.

That’s the 12-word definition of entrepreneurship that they teach at Harvard Business School.

I first read it while researching my 2010 book, The Intelligent Entrepreneur

I remember staring at it on the page and feeling like a boy noticing girls for the first time: There’s something really interesting here, but I know there’s a lot more to it than I currently understand.

I’d like to break the definition down for you, because it not only gives insight into why people like you are so drawn to the idea of starting and building something, it will also improve the likelihood that you’ll be successful.

(As a quick aside, seeing that definition in another of my books is what originally led me to meet Inc.’s editor-in-chief, Eric Schurenberg. A column he wrote about it became the most-read article in the history of Inc.com at that time.)

1. “Entrepreneurship…”

Let’s start with the word itself: Entrepreneurship. A noun with few true synonyms. (that lack of real synonyms can be a real pain in the neck.) It’s not simply a matter of being a boss or a leader or owning a business. In fact, there’s nothing intrinsic at all in this definition about business, or risk, or even making money. It’s something different–a way of looking at the world.

2. “…is the pursuit of opportunity…”

There are two key words here: pursuit and opportunity.

“Pursuit” means there has to be action involved (hence, my reader-inspired decision this year to change the name of my column to Action Required). You have to have impact; you have to try to change something. Simply thinking about an idea doesn’t cut it, and neither does coasting along doing what you’ve always done.

Similarly, a true entrepreneur is always pursuing “opportunity.” That means something new, bigger, nicer, better, smarter, more useful.

it often also means pursuing the most amazing, appealing, enticing opportunities you can find.

Here’s where we really start to differentiate true entrepreneurs from everyone else.

There are a lot of good people out there running very nice businesses. However, if they’re not chasing new opportunities–if they’re coasting along, doing what they’ve always done–then maybe they’ve given up the mantle of true entrepreneurship.

3. “…without regard to resources currently controlled.”

This might just be my favorite phrase in the world. I suppose if Harvard Business School had wanted to make the definition more accessible, they could have said “regardless of” instead of “without regard to,” but no matter.

“Without regard to resources currently controlled” means it doesn’t matter how little you have at the start. It doesn’t matter that you don’t have money, or that you don’t have all the required skills, or that you don’t have a team to help you.

At the very beginning especially, reach for the stars. Don’t let the opportunities you pursue be limited by the assets you currently have. Instead, let the attractiveness of the opportunity serve as your guide.

There are so many implications of this part of the definition.

For one thing, while capital is a necessary ingredient, the truth is that all of those would-be entrepreneurs out there who blame a lack of money for their inability to get started are playing the wrong game.

there’s an advantage to not having money at the start, because that scarcity forces you to be more resourceful. It means you have to sell your ideas to others–a possibly painful exercise, but one that pays huge dividends in the long run.

Here’s the bottom line: For just about any decision you have to make in life, there are two ways to make choices.

Most people choose the first method of decision making. They look at the array of options that seem reasonably attainable, and then pick the best one. They choose a career because it’s what their parents advised, or because there are jobs available. They live somewhere because it’s what they’re familiar with. They surround themselves with the kinds of people they’ve always known.

The true entrepreneur, however, sees things differently.

Instead of choosing the best available option, he or she thinks big, and tries to identify the best possible solution, regardless of whether it seems completely implausible and unattainable. Then, he or she gets to work, trying to make that impossible dream a reality.

If you choose the first path, you might save yourself a lot of heartache, and a lot of ups and downs on the roller coaster of life. However, you also run a greater risk of achieving your goals only to find you didn’t push yourself enough. Which path will you choose?

BILL MURPHY JR. is a journalist, ghostwriter, and entrepreneur. He is the author ofBreakthrough Entrepreneurship (with Jon Burgstone) and is a former reporter for The Washington Post. @BillMurphyJr

History Of Graphic Design: In Icons

We know two things for sure about the guys over at Brooklyn’s Pop Chart Lab: they love drinking, and they love good graphic design.

Pascal Zoghbi posted this link on May 3, 2014 via FB:
The History Of Graphic Design, In Icons http://t.co/f37hgSLUEx

Their latest poster is a tribute to the entire history of the latter: The gridded, black-and-white poster is a cheat sheet to the history of graphic design, beginning with the Victorian era.

Start at the top, left-hand corner, of A Stylistic Survey of Graphic Design, and read from left to right.

Each era (say, Arts & Crafts or Art Nouveau) is represented by a rectangular box that includes several squares that graphically represent the style described.

The Modern movement, one of the largest movements depicted here, includes Bauhaus, Vorticism, De Stijl, New Typography and Istotope, Constructivism, Suprematicsm, and Futurism.

Pop Chart creates, within each stamp-sized box, a visual representation of that particular style, with the design elements that prevailed at the time.

So the Constructivism box echoes the intense Soviet Party posters from the 1920s, the Futurism box has a bold, attention-grabbing arrow on it, and so on.

It’s telling that certain eras–eras that were niche or short-lived, or which are still emerging–get just one box. (This includes Dada, Digital, and Street Art/Guerrilla.)

Scan down to the bottom for a sampling of today’s reigning design philosophies. Are they right?

There’s data visualization, there’s the twee, chalkboard-loving school of handcrafted, and there’s flat design.

But where’s skeuomorphism?

Each box is efficiently packed, providing an at-a-glance answer to any designer who might ask: What, again, were the defining elements of the Late Modern Polish School era? For the rest of us, it’s just nice to look at.

Pre-order A Stylistic Survey of Graphic Design for an early bird price of $23, here.

[Image: Courtesy of Pop Chart Labs]

MARGARET RHODES

Margaret Rhodes is an associate editor for Fast Company magazine, where she produces Wanted …


adonis49

adonis49

adonis49

January 2018
M T W T F S S
« Dec    
1234567
891011121314
15161718192021
22232425262728
293031  

Blog Stats

  • 1,054,333 hits

Enter your email address to subscribe to this blog and receive notifications of new posts by email.adonisbouh@gmail.com

Join 533 other followers

%d bloggers like this: