Skip to main content

Project Bow's FAQ: Why is it called a lexigram when it looks like a word?

  • Author:
  • Updated date:

At this point in his development, Bow no longer needs lexigrams. He is literate and he spells out his own words. Bow developed literacy at the age of five and half years. But before that, we used lexigrams. And our lexigrams looked kind of like this:

Lexigram for the English word "banana"

Lexigram for the English word "banana"

Now, you might be thinking: "Hey, that's not a lexigram. That's just a word." Well, if you were thinking that, don't feel bad. A lot of people, including renowned primatologists and linguists, had the same reaction when they were first introduced to this concept.

So the question is: what is a lexigram and how do we recognize one?

A lexigram is a symbol that stands for a word. Regardless of how the symbol is drawn or how many subcomponents we can potentially divide it into, the symbol stands as a whole for the word it represents. It stands for the sound of the word when spoken, the meaning of the word in conversation and the grammatical part that the word can play when used in a phrase or sentence. It is a picture that stands for a word.

Is the photo for the English word "banana" that I've posted above a "spelled-out" word or a holistic lexigram? The answer depends on who you are. Can you read English? If so, you probably know how to sound out the word as well as how to pronounce it and what it means. You realize that the way the word is composed of subparts is not entirely arbitrary and unmotivated. You know that there is a relationship between the sounds that the English word "banana" is composed of when spoken and the sequence of letters that make up its written representation. But if you are someone who has only been taught that this symbol in the box stands for the English word "banana", without knowing anything more about it, the way Bow once was, then for you, it is a lexigram.

Like beauty, whether or not something is a lexigram is in the eye of the beholder.

Some Examples of Bow's Lexigrams in Both Hebrew and English


Why use lexigrams?

Chimpanzees cannot produce speech that humans can comprehend. It's an articulatory problem that has nothing to do with innate language ability. A human with a malformed vocal tract or a missing larynx would have the same problem. There are many methods of communicating without vocal speech: sign language and lexigrams chief among them.

When choosing sign language, one can select a particular sign language used by the deaf community, such as ASL, or finger spelling, as was used in the case of Helen Keller, or one can decide to make up one's own sign language especially for the child or chimpanzee in question. Many isolated families with deaf children started out using their own special system of signs, called home sign. There are many disadvantages to this choice, not the least of which is the fact the child remains isolated and can only communicate with family members. Speakers of a standard sign language can speak with the deaf community, but they must learn spoken languages as if they were a foreign language. Fluent speakers of ASL often write English as if they were foreigners, struggling with syntax and grammar and word choice, even though they may have grown up around native English speakers all their life. It isn't just a question of pronunciation. ASL is a completely different language from English, with its own distinct grammar.

In choosing lexigrams for Bow, I was following in the footsteps of Duane Savage-Rumbaugh and Sue Savage-Rumbaugh. But while their use of lexigrams was more like "home sign", I wanted Bow to be able to communicate with the entire world.

For this reason, I chose lexigrams in three standard human languages. Our lexigrams consisted of words written in the standard orthography of the language in question. The three languages were Hebrew, English and Mandarin Chinese. You can see examples of our earliest lexigrams here.

Eventually, because we did not have enough volunteers who spoke Chinese fluently, we had to drop Chinese from our list of languages at Project Bow. Bow grew up bilingual, speaking Hebrew at home and English with outsiders.

A Brief History of Lexigrams

Wouldn't it have been easier to use iconic pictures for our lexigrams? Wouldn't that have spared Bow the trouble of having to deal with the arbitrariness of the orthographies in English and Hebrew? The answer is a resounding no! However, in order to understand why, it would help to know a little about the history of lexigrams, both in ape language studies and in the ancient world.

In the 1970s Duane Savage-Rumbaugh worked with a female chimpanzee named Lana using plastic lexigrams and a made-up language called Yerkish. Lana was two and a half years old when she started the project, and she helped to demonstrate that chimpanzees were capable of manipulating symbols. This was very important pioneering work in the field of ape language studies. However, because Yerkish was not a standard language, nobody was a native speaker of Yerkish, and many issues involving natural language acquisition and grammatical usage could not be tested. Likewise, Lana did not have the opportunity to use Yerkish with a broad spectrum of interlocutors. Many things about the experiment were unnatural, and so the highest language potential of chimpanzees could not be demonstrated.

Sue Savage-Rumbaugh went further with lexigrams in her work with the bonobos Matata, Kanzi and Panbanisha and their growing family. This time, there was no artificial, made-up language involved. The bonobos were exposed to spoken English, and the lexigrams they used stood for English words.

However, the lexigrams that were used emphasized picture recognition, and even when they sometimes contained English words in standard orthography as part of the picture, the approach was to make the color and the font and the general appearance more important than the letters.

A keyboard with Some of Kanzi's Lexigrams

Image Credit: The Great Ape Trust Website

Image Credit: The Great Ape Trust Website

From Trademarks to Lexigrams

If you look at the lexigram keyboard above, used by Sue Savage-Rumbaugh with Kanzi, you can see that some of the lexigrams are purely pictorial, while others embed recognizable Chinese characters or English words as part of the picture. But the standard orthography that is embedded in the lexigram is stylized so that its font, color and background are equally as important as the letters or characters embedded in the picture.

Take, for instance, the lexigram for Perrier water. It looks very much like the Perrier trademark. It includes the letters P-e-r-r-i-er, but as with all trademarks, that isn't the most salient feature. There is the green circle and the font and the overall look of the thing to contend with.

Scroll to Continue

The emphasis on non-alphabetic cues has the effect of hiding the alphabet in plain sight, making it less accessible to the subject. This is not a good thing, for grammatical reasons as well as for purposes of communicating with the outer world.

Proto-Semitic Aleph

Image Credit: Wikipedia Proto-Semitic Aleph

Image Credit: Wikipedia Proto-Semitic Aleph

Here is an experiment you can try at home, if you are an English speaker. Download Kanzi's keyboard for yourself -- or create your own lexigraphic keyboard -- and try to communicate with others in English using only the words for which you have lexigrams. Here is what you will find: not only will you be extremely limited in your vocabulary, but you will also find yourself using a kind of pidgin English. Why? Because English has inflection, and in order to speak grammatically, you have to use it. Even if you have special lexigrams for -ing or -ed or -s, how are you going to handle see/saw or eat/ate/eaten?

Other languages, such as Hebrew or Latin or Turkish have even more grammatical morphology than English. This is why, long, long ago, the ancient Semites went from using lexigrams to using symbols to stand for speech sounds. That is in fact how much of mankind came to use the same alphabet. The Greeks borrowed it from the Phoenicians. The Romans borrowed it from the Greeks. The English that you and I use on Hubpages is written in letters borrowed from the Romans. Did you know that the letter "A" derives from semitic Aleph, and that it used to be a picture of the head of a bull?

From Lexigram to Letter: The History of Aleph

The Phoenician letter aleph was originally a lexigram for  the word aleph "bull"  Image Credit: Wikipedia

The Phoenician letter aleph was originally a lexigram for the word aleph "bull" Image Credit: Wikipedia

The alphabet is a great invention, but many people assume that it requires a great deal from writers and readers, and that acquiring literacy is a sign of some kind of major intellectual achievement.

In fact, alphabetic writing requires less of speakers than does lexigraphic writing. The great achievement of our current writing system is its simplicity and the way it mimics what we are already doing when we speak.

Instead of memorizing long, long lists of words and their pictorial representations, an alphabetic writing system allows you to write any word you can say, without special training. (This is less true in English than it is in Hebrew, because the English writing system is less phonemic, but that is a different story for a different hub.)

Roman A: Not an Original Invention

Image Credit: Wikipedia The Roman A is borrowed from Greek Alpha which was borrowed from Phoenician Aleph

Image Credit: Wikipedia The Roman A is borrowed from Greek Alpha which was borrowed from Phoenician Aleph

One of the claims made against chimpanzees and bonobos who use lexigrams is that they do not speak grammatically. This is not because they are not human. It is a function of the mismatch between lexigrams and the inflectional language that they are supposed to model. Anyone would have the same problem when limited to this means of communication.

As long as Bow had to limit himself to lexigrams, he didn't appear to have grammar, either. But the moment he started spelling out words, that problem completely disappeared.

We didn't teach Bow how to read. He figured it out spontaneously, much as a baby learns to understand speech. When a baby learns language, he does it by attending to the speech in his environment.  At first the words and sentences seem like indivisible wholes. Then, gradually the child discovers phonology. He realizes that words are made up of a limited inventory of vowels and consonants. He unlocks the secret to language.

But isn't reading really hard? Not really. What's hard is not having access to the subcomponents of words and still being expected to use language productively. In a sense, to keep the alphabet hidden from a chimp because you think he can't make it out is like deciding to use holistic pre-verbal cries while conversing with your infant, because he hasn't yet acquired phonology. He'll never acquire it, unless you give him a chance. Exposure and immersion is key.


One of the contributions of Project Bow to ape language studies is the use of lexigrams composed entirely of words in standard orthography. By doing this right from the beginning, we allowed Bow to discover the alphabet for himself.

(c) 2010 Aya Katz

When Sword Met Bow


Aya Katz (author) from The Ozarks on May 05, 2011:

Yes, I may not know exactly how he learned. This is not an experiment focused on pedagogical methods. It does focus on giving the subject the opportunity to access sub-components of words, which is really what this particular hub is about.

If you read this article, rather than some other article by me about Project Bow, then you see that when subjects are not given access to the phonological or at least the morphological level of language (whether visually or otherwise) then they cannot construct well formed sentences.

I believe this is a contribution to science that I have made independent of the success of Project Bow as a whole.

skepticality on May 05, 2011:

re my commenton the TV clip aboutSue: On this specific point I was criticising only themoderator who stated thatKanzi taught himself to usehuman language. The clip had nothing todo withthat, but dealt with some test that involved digits on a computer-screen. (Incidentally, it was for that test that Sue put a mask on her face). But, as for the moderator statement about Kanzi teaching himself to use human language, the circumstances under which the statement wasmade misled meto assume that Kanzi learned to use a human language just by hearing and seeing humans talking; which was not at all the case!

I brought up this issue, because sue was obviously trying to train Kanzi's "mother" to use a human language (although I am not sure how she did it; the article didn't explain). But she was not doing the same thing with Kanzi. I simply wanted to pointout thatyou can deliberately train an animal like Kanzi (or Bow), to do certain things, and that they may also learn by just being given the opportunity, even quite accidentally, and unknowingly.

The problem I see with Bow's aculturation is that, it is quite possible that in many cases, you may not even know what you gave him an opportunity to learn.

Aya Katz (author) from The Ozarks on May 05, 2011:

Skepticality, you misunderstand the nature of the experiment. I did not train Bow to do anything -- if you exclude potty training. ;->. I exposed him to a great many experiences, and he learned what he learned by himself. I did not teach him literacy. He picked it up. But I gave him the opportunity.

I find your criticism of Sue Savage-Rumbaugh to be based on a false assumption: that if Kanzi attended to Sue's instruction of Matata, that his learning was not spontaneous.

All of us learn only from experiences we have been exposed to. None of us invent language. We need the exposure. But on the other hand, you cannot teach language by rote. The learning that occurs is in the mind of the individual, and it's what he makes of his experiences that creates his store of knowledge.

It is impossible to document all of Bow's experiences. If proof eventually emerges, it will be of what he can do on his own -- but not of how he came to be able to do it.

skepticality on May 05, 2011:

You seem to me to confuse the facts, i.e. your results (data) with your interpretation of your data. Ther facts need no proof, and cannot be questioned; unless you have reason to suspect that something went wrong during your test; which requires that you repeat the tes more than just once.

As for Bow's aculteration, I see a serious problem here, because by bringing him up as a human child, you constantly exposed him to an almost endless number of experiences, without being able to even jot down all these experiences, and keep a fully detailed log of them.

This means that your study is being done under uncontrolled conditions. You know what you deliberately trained Bow to do. But you do not know what he might have learned just because you gave him the opportunity to learn through his experiences (of course, depending on what, as a chimp, he had the ability to learn; which you do not know).

This reminds me of a TV clip about Sue , where the moderator stated that Kanzi taught himself to use human language. But when I read a popular article about that, it turned out that he was always present while Sue was, unsuccessfully, attempting to teach the use of human language to a five years old female (that treated Kanzi as her baby). Sue said Kanzi usually sat on her (Sue's) head, and paid no attention to what was going on. Obviously, even though she did not train him to use human language, she gave him the opportunity to learn this by hearing and seeing what was going on with the training of his "mother". If he seemed not to pay attention, it is possible that Sue had no idea what a bonobo should look like when he is paying attention; or else, he paid sufficient attention to learn what he could learn, and stopped paying any further attention, when Sue repeated, over and over again, her unsuccessful attempts to teach his "mother".

Aya Katz (author) from The Ozarks on May 05, 2011:

I think it is important to distinguish the research project from the "proof". Even before Bow began to spell, I considered that I was involved in a research project, when all I was doing was taking care of him as I would any baby. This was laying the groundwork for his enculturation in our society and our family. Without this, the experiment would not exist.

I understand that I am not proving anything right now. But I would not say that the only way to continue the research is to work directly on rigorous proof. The fact that the enculturation continues opens the way to better proof further down the line. If I were instead to destroy my relationship with Bow in an attempt to obtain proof now or never, then it would indeed be the end of everything.

I see it this way: there are the facts, and then there is proving the facts to the satisfaction of others. The facts come first. The proof may come later. If it doesn't come, that's okay, too.

This is an attitude that would really have helped someone like Herbert Terrace to put his research in perspective.

skepticality on May 04, 2011:

You continue to talk to Bow, while he responds by typing on the glass. Under such circumstances you cannot exclude cueing (due to the "Clever Hans" effects),by body-language, facial-expressions, and tactual-contacts that you must maintain with Bow when he types.

Perhaps you consider this a way to continue your research. I consider this as merely continuing to use improper reserarch-methods that you already know you should avoid using.

Aya Katz (author) from The Ozarks on May 04, 2011:

There is no deadline for the research project. Bow and I talk every day. He spells spontaneously on the glass. He communicates with Lawrence, his sitter, also by spelling out words. When and if he decides to cooperate by letting others see this, we will make that public. If this never happens, that is also all right. We have all the time in the world, and nothing has been discontinued.

With children, also, they sometimes go through a stage when they only talk to family members and people they really like. When a child is being selectively mute, you can't force him to talk to others, and sometimes outsiders do not believe he can talk. (Look up selective mutism). Force doesn't work. Patience and understanding work.

Irene Pepperberg did not need to wear a mask with Alex and there was no doubt that he chose what to say. As long as we don't see the subject selecting the subcomponents of words, then we cannot cue him. That is an important point to keep in mind.

skepticality on May 04, 2011:

I did not ask whether you will dump Bow, but only whether you will dump your research project on him

Your answer to what you now intend to do about it was that there's nothing you can do.

Do you mean you will not try to ask him what he wants to eat and leave him only the touchscreen to respond, because you fear the possible results? Do you fear he might go on a hunger-strike to the death? Attack someone? Or what?

Is there any chance you could return to the situation you had at the start, when Bow used the "glass" to type, with no one touching him, but typed too fast to see what he was typing? If you could, you might then be able to videotape his typing, and slow down the videtape, instead of trying to slow Bow down. And, of course, you would have to wear a mask on your face, (as Sue Savage-Rambau does with her bonobo), to exlude cueing by facial-expressions, or by where you are aiming your gaze. She does nothing to exclude cues by other body-language, which I see as a problem. And I doubt you could do anything about this problem, because it appears that Bow needs to see you to try to communicate with you.

Aya Katz (author) from The Ozarks on May 03, 2011:

If we are recording the session, then the video/audio will confirm what he says, and many "witnesses" who are not there at the time can check our interpretation of what the computer said. It's not a serious problem. The Clever Hans accustion will be circumvented, because we could not have cued the letters, and the accustion of not hearing correctly will be something that disinterested parties can check from the record.

As for the fact that Bow is not cooperating, there is _nothing_ I can do about that! Will I dump him if he doesn't do what I want? Of course, not. I am committed to Bow, because I adopted him. He's mine for keeps, language or no language.

Thankfully for me, for Bow and for Sword, I have a few other talents that I can use to make a living and support us all.

This means, for purposes of our family, that we will survive.

What it means for the purposes of science is that we are in this for the very long haul. It's bad science that demands instant results. It is dirty science that depends on funding for its conclusions.

skepticality on May 03, 2011:

Your only chance to completely exclude the possibility of cueing Bow when he types, is to get him to use the touchscreen. When someone like Lawrence was not close enough to see what Bow was doing, but close enough to hear the "talking" computer, it would seem to me that the only way for Lawrence to be sure he heard with perferct accuiracy, would have been to check what was printed on the computer-screen. You say that this was not always possible, because you used the computer also for playing games. In that case, the least you should have done, is erase all trace of the game, once the game was over.

Now it seems, however, that you have a totally different problem, because Bow refuses to use the touchscreen altogether! So what are you planning to do about it? Are you hoping his decision is not final, and waiting for the chance that he might still change his mind, or what? Will you dump your whole research if he won't change his mind???

Aya Katz (author) from The Ozarks on April 04, 2010:

Thanks, drbj!

drbj and sherry from south Florida on April 04, 2010:

Fascinating hub and work, Aya. I'm impressed.

Looking forward to reading more.

Aya Katz (author) from The Ozarks on April 04, 2010:

Thanks, Larry. The issue of how to remain sane when Bow is complaining "She got more!" and Sword is upset that Bow has been rude to her is a whole other story! But that problem is so universal that it hardly rates a hub. ;->

Larry Conners from Northern Arizona on April 04, 2010:

Aya...This is an extremely interesting and well-written Hub on a subject I am blissfully unaware...I can't imagine the trial and error methodology involved...the patience and creativity required to achieve measurable results...

The stress and frustrations encountered raising my two daughters cost me my hair...yours appears firmly attached, so I assume you take it in stride as you develop Bow's communication skills...

Thank you for sharing this most interesting read...Larry

Related Articles