The Experiment...

by Margaret Magnus

copyright by Margaret Magnus
all rights reserved

Margo's Magical Letter Page

I would like to encourage people who read this page to try the following simple experiment. In about 3 hours, any native speaker of English can convince themselves beyond a reasonable doubt that the relationship between the sound of a word and its meaning is not arbitrary. One of the great mysteries for me in this life is why even most professional linguists do not in general acknowledge the extent of the correlation between sound and meaning given that the proof is so simple, and given that it does not require $12,000,000 in lab equipment, but merely a pen, a piece of paper, and a collegiate sized dictionary. Furthermore, the consequences of this observation for our understanding of language and thought systems in general seems to me vast. It has all kinds of practical applications in areas like machine translation and search engines and foreign language learning. And what's more important for the impractically-minded hermits like myself, it provides fundamental insights into the workings of the human mind.

Experiment ­ Part 1: Pick your favorite consonant in a language you speak fluently. This consonant is to be thought of as a sound, not a written letter... so the /k/ sound in 'kite' and 'cream' is the same consonant even though they're spelled differently, and the 'c' in 'caviar' and 'ceiling' are different consonants, even though they're spelled the same. Open your dictionary to a section containing words that begin with that consonant. Find a monosyllabic word in the dictionary beginning with that consonant. (If the language you are using does not have a large supply of monosyllabic words, simply use roots.) Write it on your sheet of paper. (Or preferably use a word processor.) Find another monosyllabic word beginning with that consonant. If it is very similar to the first word in some way, write it on the same line as your first word. Otherwise skip down a few lines and put the second word on a new line. Find yet another monosyllabic word beginning with this consonant. If it is very similar to either of the first two words, then put it on the same line. Otherwise put it on a new line. Continue to do this for about 2 hours with as many monosyllabic words in the dictionary as you can. All the words on a single line, or in a single class, should have some single fairly narrow element of meaning in common and should begin with the consonant you have chosen. How narrow should this element of meaning be? Remember that the purpose of this experiment is to determine for yourself whether my claims are true. So the class should be narrow enough that it seems to you to be highly unlikely that the semantic similarity of these two words is mere coincidence. I generally begin by combining words into a single class which are about as closely related in meaning as 'flame' and 'fire', or 'bloat' and 'bulge' or 'slide' and 'slither'. If you know something about etymology, forget it. Don't assume you know where the words came from.

Result ­ Part 1: Within two hours, you will find that about 70% of those monosyllabic words fit in a very few very narrowly defined and related classes. Each word will fit on average in about 2 classes. If you give the experiment 20 hours, 95% of the words will fit in a few very narrowly defined classes slightly reorganized, and the average word will fit in about 3.5 classes. If you entrust your life to this and related experiments, every word will reveal itself to be influenced by its sound in very specific ways.

Experiment ­ Part 2: But no, no! Don't go away! We're not finished yet. You promised me three hours of your time, remember? Not just two. With the remaining hour, pick another consonant that you also especially like. Now continue the experiment on the same sheet of paper. Try putting the words beginning with, for example, /n/ into the classes you've just defined for /k/ (or whatever your consonant was).

Result ­ Part 2: The new consonant will do about 1/2 to 2/3 as well as the old consonant depending on how similar they are. Over time, however, words beginning with the new consonant will reveal themselves to fit differently. All the 'puffy' /p/ words that start falling into the 'bulging' /b/ class will differ from the 'bulging' /b/ words in some fairly narrow and well-defined way. In this particular case, the bulging /b/ words involve a membrane or surface which is filled with liquid or air under pressure, and the /p/ words will be more textured, and will not necessarily involve any membrane or air or water. My guess is that by the time you have completed this much, you will at least be seriously considering the validity of the phonosemantic claim.

Experiment for the Die Hard: So if you find yourself still willing to indulge me, and if you speak a second language, take those same two consonants, if they exist, from the other language and continue on the same sheet of paper. Tell me now if you original consonant doesn't work more or less the same in both languages.

I have also spent years wondering nearly to the point of distraction how it is still possible 2400 years (2400 years!!!!) after Plato so neatly described the abovementioned experiment in his Cratylus dialogue, that we still believe the claim that the relationship between sound and meaning is arbitrary. By now the only hope seems to me to appeal to you by means of this marvellous technology known as the Internet.

For more on:
Margo's Theory of Phonosemantics

Margo's Magical Letter Page