Was Saussure wrong?

Welcome to the second installment in our series on Ferdinand de Saussure and the linguistic science of semiology. Now where were we?

In the last post we discussed Saussure’s theory of the “sign” as a combination of the “signified” (the concept represented by a word) and the “signifier” (the spoken or written word doing the representing). According to Saussure, the relationship between the concept and written/spoken word is arbitrary. The word doesn’t matter as long as it’s different from every other word in a given language and every speaker of that language accepts that it represents the same thing.

Despite a wealth of evidence in its favor, the distance between signified and signifier doesn’t sit well with a lot of people. Is it an attack on our personal relationships with our language? Or, worse, does it attempt to prove that those relationships never existed?

If this dichotomy worries you, you’re not alone. These issues were on Plato’s mind over 2000 years ago. In his dialogue The Cratylus the philosopher investigates the differences between linguistic “conventionalism” (akin to Saussure’s theory that words have no inherent tie to the concepts they represent) and linguistic “naturalism” (the lovely but scantily supported view that words naturally belong to the concepts for which they stand). Plato includes Socrates as a character in the dialogue to moderate the discussion and play devil’s advocate by invoking wild historical etymologies. True to form, Socrates does his job a little too well; by the end of the dialogue, the issue is more complicated than when the discussion began.

The question of this fundamental relationship floated around philosophical circles long after The Cratylus, most often with uncertain conclusions, only to reemerge in the literary realm. Shakespeare’s “A rose by any other name would smell as sweet” supports a conventionalist view in Romeo and Juliet. In Through the Looking Glass, Lewis Carroll nods to the naturalist view when Humpty Dumpty declares, “My name means the shape I am,” implying that the sound of “Humpty Dumpty” evokes an image of a round egg-like figure. (See the piece from The New Scientist on this impact.)

In spite of this ancient debate, neither philosophy nor literature could threaten Saussure’s hypothesis…thank goodness science stepped in.

Meet Bouba and Kiki.

Look at the two shapes pictured here Which would you label “Bouba” and which would you label “Kiki”? Go on. Guess.

Neuroscientists Vilayanur S. Ramachandran and Edward Hubbard developed this experiment in 2001, a reworking of an earlier study by psychologist Wolfgang Köhler. The goal was to investigate the ability of the human brain to derive abstract properties from shapes and sounds. The experiment tests for properties of synesthesia, a neurological condition in which stimulation of one cognitive pathway involuntarily fires another. A synesthetic might assign certain colors to letters and numbers, or associate various sounds with physical feelings, but the Bouba/Kiki experiment was the first to detect synesthetic tendencies within language.

Back to those shapes: If you called the rounded shape “Bouba” and the spiky shape “Kiki,” then you can count yourself among the vast majority. Ramachandran and Hubbard tested English, French and Tamil speakers and 95 to 98 percent had the same associations. The results of this study represent the first scientific retort to Saussure’s hypothesis of the arbitrary relationship between the signified and the signifier. (Here’s the whole study.)

What sounds (and words) do English speakers think are the grossest? Find out here.

Perhaps supporters of the naturalist hypothesis were just asking the wrong question. Why reach back into the annals of our etymological history to try to link the signified to the signifier, when we can find a common tie in our present-day linguistic instincts.

We’ve just presented a rather rosy rebuttal to Saussure’s argument, but you can bet he has something to say about it. Here’s the shocking conclusion to this tale of sighs and signs.


  1. […] “Back to those shapes: If you called the rounded shape “Bouba” and the spiky shape “Kiki,” then you can count yourself among the vast majority. Ramachandran and Hubbard tested English, French and Tamil speakers and 95 to 98 percent had the same associations. The results of this study represent the first scientific retort to Saussure’s hypothesis of the arbitrary relationship between the signified and the signifier. (Here’s the whole study in Synaesthesia in pdf format – very interesting.) Info from Blog Dictionary.com“ […]

  2. Lydia -  May 15, 2013 - 1:02 pm

    I wasn’t in the majority. I saw this:
    Kiki sounds like kiwi, and kiwis are rounded. Hence i saw the rounded shape as Kiki
    Bouba reminded me of the word “bo” (with a dot under the “o”) in Vietnamese, which has a sharp sound to it. So therefore I saw Bouba as the sahrper shape.

  3. Scott -  May 14, 2013 - 5:52 pm

    Saussure is pronounced So-sure? My cats is sleeping all over the key-board, can’t comment.

  4. Bendrix -  April 9, 2013 - 6:43 am

    I’ll believe it once they expand the base of language speakers being tested. Two speak European languages, and one speaks a Dravidian language. Add in some Sino-Tibetan, Arabic, African, etc. speakers and this research will be far more convincing.

  5. Kay -  February 15, 2013 - 7:53 pm

    Maybe it has to do with the sound. Or maybe that’s only my opinion. To me, just pronouncing a the letter “B” has a smooth feeling in my mouth, whereas pronouncing “K” has a scratchy and rough feeling when repeated several times. The same applies to an “A” versus an “I”.

  6. minde28383 -  February 14, 2013 - 7:41 am

    my brain worked like this for words: Bouba and Kiki
    Instantly i made association Bouba – bubble ie. rounded shape
    Kiki – kick ie not rounded

    that’s how i did my guess.

  7. Maya Pinion -  February 1, 2013 - 4:19 pm

    A child would call “kiki” a “star” shape and “bouba” a “flower” shape if asked simply, “What is it?”. Language is much more than a sound or shape association due to synesthesia. It is somewhat random, somewhat logical, historical and borrows from other languages. It doesn’t normally follow the association rules that synesthesia may naturally create. There are only 26 letters in English which must create words that represent our ideas, feelings, thoughts and our real world. This is limited considering the potential of the human brain. Have you ever searched for a word which you discover later does not exist? You could create your own word, but what rules would your word follow? Probably logical rules, borrowing from other words, prefixes and suffixes. People do this all the time when creating a word like, say, “uncomfortableness” instead of “discomfort”. To a child who does not know the language, it seems correct. But of course, we grow to accept anomolies, rules, and limits. The answer to the question, “Was Saussure wrong? is yes and no. Many words may have begun through a synesthetic process, but later were transformed beyond recognition.

  8. Abby -  January 29, 2013 - 2:17 pm

    I love this series of articles. These are very interesting theories. I am in an acting class, and we are performing Alice in Wonderland/Through the Looking Glass. I play Humpty Dumpty, and when I read this article I wanted to add something. At least in our production, after this line, Alice introduces herself and Humpty Dumpty says, “Why, with a name like yours, you could be any shape, almost.” Is it possible that the theory in this article is only true for SOME words?

  9. Dilia Bobadilla -  January 28, 2013 - 5:46 pm

    “At the end of the day” the bubba kiki test is yet a subtle argument with no clear explanation of how associations happen to form a consensus or majority. Btw, I am not in the majority. I’m not sure at what point neurology picks up or if it’s even inherent or exclusive to the association itself.

  10. Megan -  January 28, 2013 - 6:32 am

    Imagine instead of “bouba” and “kiki” one was “boki” and the other was “kibo”. What would your impressions be as to which figure is which?

  11. Bruce -  January 27, 2013 - 5:44 pm

    And I can’t wait for Segment 3!

  12. Bruce -  January 27, 2013 - 5:44 pm

    I think both sides are correct in this. We generally use sound signs that are arbitrary, but we may have cultural and even physical influences that we associate with emotion or meaning.

  13. mary -  January 27, 2013 - 11:46 am

    I agree with @Megan. I think that the shape of the letters, and the taste of the word in your mouth definitely make a difference.

  14. rampartwatcher -  January 27, 2013 - 7:37 am

    Your point, Musicspren, is very relevant inasmuch as the 2 of the languages in question are in the Indo-European “family” and the Dravidian dialect of Tamil is an directly adjacent (Southern India primarily) “tribe” (which may be the antecedent thereof), so the vocalization and aural recognition of such “signs” may be similar enough to establish similar neural encoding to support the obvious synesthetic effect. Would such occur in completely different language groups that may be tonal in interpretation (some Asiatic) or even truly onomatopeic ones (some African) is the question. As a totally amateur ethno-linguist, I look forward to the next post (and perhaps a subsequent series on semantics).

  15. Mike -  January 27, 2013 - 7:09 am

    Where did my comment go? The concepts of leading and arguing are rather like the sounds bouba and kiki. They seem like opposites. They start before words in the history of a human learning.

  16. Mike -  January 27, 2013 - 7:00 am

    One thing leads to another – another truth is one thing fights with another

  17. Eric -  January 26, 2013 - 4:05 am

    Nice to see a preponderance of intelligent comments for a change!

  18. SAUSSURE-2 | BLOGCHI@mayopia.com -  January 25, 2013 - 7:49 am

    [...] “Saussure-2” — Oh what to do — Technology so rapidly changing — The Universe of Communication — The Context of every situation, — Revealing rearranging. — Every Rule is eventually broken — With each and every new dialect spoken — Any almost study breaking — ‘Methods‘. — Funding for the Faking. — Monetary wrangling. — Scientifically mistaking — A “Kiki” image for a “Bouba” token. — Simply out of date and dangling. –>>L.T.Rhyme This entry was posted in DICTCOMHOTWORD, L.T.Rhyme, RLTR and tagged LT, LTRhyme, the HOT WORD on January 25, 2013 by LTRhyme. [...]

  19. steve -  January 25, 2013 - 6:27 am

    I’m suspicious of the question.

    Naturalists are certainly on to something. But “bouba” and “kiki” seem to have more to do with word origins (or improvised-noises-that-sound-like-words) than how we actually talk.

    Ol’ Saucypants seems to be talking about how we, as trained speakers, actually use language. So really, it’s apples and oranges. (Words with colors?)

  20. Joe -  January 24, 2013 - 3:40 pm

    Maybe related to morphemes?

  21. Zeke -  January 24, 2013 - 8:19 am

    (What sounds (and words) do English speakers think are the grossest?

    Why is this sentence even include here? You took what was actually an interesting article, and cheapened it to the level of a fart joke in one quick stroke.

    Don’t get it…

  22. Verbie -  January 24, 2013 - 6:42 am

    No doubt that languages are composed of varying types of sounds, some abrupt, like the sound of ‘K’, others not so abrupt. Besides onomatopoeias, some words seem to have a connection between meaning and sound, like “cacophony” or “azure” — but this would likely happen often even if the assignment of signifier to the concept were random or arbitrary.

  23. Bubba -  January 24, 2013 - 6:41 am

    I have always connected numbers with colors. Not so much any more (I’m 60) but as a child the colors were Sesame Street bright. Even the flavour of certain things would evoke color associations. “mmm, that tastes GREEN!) This taste thing was likely learned but nonetheless vivid in my four year old brain.

  24. Bubba -  January 24, 2013 - 6:25 am

    I wonder if being written in English might be more than somewhat leading. Bouba has an obvious ovate shape while the letters we use to spell Kiki are spikey and sharp shaped. Is there connection between shapes of letters and the sounds they represent? How about the shape your mouth makes when forming a sound (OH and SPIKE) and the shape of the letters used? It seems that there are obvious connects and it would seem a natural development/evolution. I think that on a basic tribal level, primitive language would tend to be somewhat organic in texture and as vocabularies grew, more and more exceptions would come to be included.

  25. Megan -  January 24, 2013 - 1:56 am

    Perhaps it has something to do with the shape of the letters used in the words? B, O, U, being rather rounded, K and I being more sharp, so that one associates Bouba to the form with rounded ends, and Kiki to the pointy form?

  26. Musicspren -  January 23, 2013 - 3:20 pm

    I’m suspicious of “Bouba” and “Kiki” with English speakers – we have words such as “bulbous” and, as this mentioned, “spiky.” If they tested only three languages, one of which was English, it could easily overbalance by similarity to other words, not anything inherent (however, I know neither French nor Tamil, so it’s possible they don’t have the same similarities).


Leave A Comment

Your email address will not be published. Required fields are marked (required):

Related articles

Back to Top