From 4 to 6 July the second Minding Animals Conference was organized at Utrecht University. Every three years this conference is held in another university city, set up by Minding Animals International. The purpose of these conferences is to build bridges and bring people together from science, philosophy and animal organizations. During the conference many lectures are given about all kinds of subjects that have to do with animals: the intelligence or cognition of animals, emotions and consciousness of animals, animal ethics, politics and policies regarding animals, animals in literature and art, and yet more subjects. Hundreds of people from all over the world had travelled to Utrecht for this 3-day conference. The programme was packed and started each day at 9 in the morning, to end only at 10:30 in the evening. It was a tiring but very interesting conference.
Christian democrat minister of agriculture Henk Bleker opened the conference with a long self-apology about his farming background, assuring all present that farmers also have a good relationship with their animals.
My two lectures were scheduled for the first day, Wednesday July 4. In the morning during the session “Animal capacities: concepts, beliefs and language” I gave my lecture Have nonhuman great apes acquired human language? In this lecture I gave a review of all language research with nonhuman great apes that has been carried out for over a century now. I discussed the results of these studies in which great apes were taught some form of human language, by spoken language experiments, sign language projects and communication by geometric symbols or lexigrams. Below you will find the text of both lectures. The conclusion of my lecture was that chimpanzees, bonobos, gorillas and orangutans are able to learn individual symbols for all kinds of objects and actions, but that nonhuman great apes do not combine these symbols in a meaningful, structured way and that they have a limited motivation for using these symbols: in order to obtain things and actions from humans. For this reason I ended the lecture with the position that nonhuman great apes have not acquired a human language and that the capacity for language is a qualitative difference between humans and other apes. In general the lecture was received well. However, some people were not pleased by my conclusion. The fact that I don’t ascribe language to other animals amounts to some people to confirming or creating anew a dichotomy between humans and other animals. Unfortunately, people then don’t listen good enough anymore to the scientific arguments that I presented.
In the afternoon I gave my second lecture during the session “Animal ethics: capacities and relations.” This lecture was called Is absence of language in nonhuman great apes morally relevant? and was the moral sequel to the scientific conclusions that I had presented in the morning. In the lecture I criticized the positions of philosophers such as R.G. Frey and Peter Carruthers, who take the capacity for language as an important moral distinction between humans and other animals. I gave my criticisms of their arguments and came to the conclusion that the capacity for sentience (the ability to experience pain and pleasure) or phenomenal consciousness (the experiential aspects of our mental states) is a sufficient condition for a moral equality among all animals. In my view the presence or absence of language does not lead to a morally different treatment of humans and other animals, and I argue for equality of all animals, with or without language. Again the lecture was visited well, among the audience the famous philosopher Peter Singer. The reactions were very positive, people thought it was a good discussion of the language issue. Some people came to me afterwards and said they thought it remarkable that I argue for cognitive or motivational differences between humans and other animals (no language), and yet that I argue at the same time for a completely egalitarian relationship between all animals. They thought I was a pleasant exception in that, which, sadly enough, is indeed the case.
Below you will find the text of both lectures. Let me know your opinion about them.
Have nonhuman great apes acquired a human language?
Before I could start the chairman of the session and me had to solve some computer problems.
When psychology and linguistics started to develop as scientific disciplines in the 19th and 20th century, a scientific oddyssey was embarked upon to discover whether nonhuman great apes could acquire something of our human language. Though several scientific questions laid at the base of this oddyssey, such as the question how much of behaviour is innate or acquired by learning, a major motivation for these studies has also been plain scientific curiosity: how far can our closest furry cousins go in acquiring parts of the human language when we teach them this?
Chimpanzees, bonobos, gorillas and orangutans have been used in all kinds of language studies for almost a century now, starting in 1909 with the American psychologist Lightner Witmer and the chimpanzee Peter. Later this type of research with nonhuman great apes has been called Ape Language Research or ALR. In this lecture I will present the results of this century of ape language research and I will answer the question whether nonhuman great apes have acquired a human language.
Before I continue, let me give you the definition of language that I use. Language, in my opinion, consists of three main characteristics. First, the use of words or signs as symbols, referring to things in the outside world or to abstract concepts. Second, the ability to meaningfully combine these symbols by means of grammar and syntax. Third, a rich variety of functions. The functions of human language can be summarized as follows: the transmission of information, expressing one’s thoughts and emotions, the regulation of social relationships, clarification of ideas, humour and play with language, lying, referring to things displaced in space and time and the telling of stories, fantasies and all kinds of abstract concepts. Language gives us the ability to express all these functions in a very detailed and specific way. Indeed, this conference itself is a good example of the possibilities of human language.
When scientists first tried to teach language to nonhuman great apes, only spoken language was considered to be a real human language. The first ape language experiments were thus focused on teaching apes to speak words, starting in 1909 with Lightner Witmer from the University of Pennsylvania, who tried to teach the 5-year old chimpanzee Peter to speak the word mama and the letter p. Later other scientists followed with their own chimpanzees and also orangutans. Usually they tried to make the ape speak a word by manipulating the mouth and lips of the ape subjects. Often the apes were living in the human homes of the scientists, where they were sometimes cross-fostered, or raised by humans as if the ape was a human child. The spoken language experiments were quite unsuccesful, however, the biggest achievement coming from the chimpanzee Viki, who lived with Keith & Catherine Hayes from 1947 to 1953. After these 6 years of training, Viki was able to produce mama, papa, cup and up. However, Viki had a lot of trouble producing these words, which were often inaudible, and more importantly, used in the wrong contexts. The failure of the spoken language experiments was explained by the differences in vocal apparatus between human and nonhuman apes: the nonhuman apes were physically unable to produce actual human words.
In the 1960s new research on human sign language showed that signed languages are just as complex and rich as spoken languages, elevating sign language from its originally prejudiced position as not a real language. This change in perception of human language caused a new chapter in ape language research: now it was the question whether nonhuman great apes could learn and acquire signs. In 1966, the psychologists Allen and Beatrix Gardner from the University of Nevada at Reno started with Project Washoe, in which they taught signs to the wild-caught chimpanzee, Washoe, who was about 10 months old when the project started. Cross-fostered by the Gardners, she was immersed in a linguistic environment in which the humans communicated by using signs. Moulding the hands of the chimpanzee into the right configuration for a sign turned out to be the most succesful method by which to teach signs. After almost 4 years the Gardners had made a historic achievement: Washoe had acquired about 132 signs for various objects and actions. This success with Washoe inspired further projects with other apes. The Gardners and later Roger and Deborah Fouts taught signs to 4 more, newborn, chimpanzees and according to the Fouts the adopted son of Washoe, Loulis, learned signs from the other chimpanzees, sometimes even by active teaching from Washoe. In 1972, the psychologist Francine Patterson of Stanford Univeristy started Project Koko, in which she taught signs to the gorilla Koko, and later to her male companion the gorilla Michael. In 1978, lastly, the anthropologist Lyn Miles from the University of Tennessee at Chattanooga started Project Chantek, with the orangutan Chantek.
Besides projects in which apes were taught sign language, since the 1970s other projects were set up in which apes were taught symbols in the form of lexigrams, geometric, often arbitrary symbols, on a plastic board or computer screen. Duane Rumbaugh started in 1971 with Project Lana, in which he taught the chimpanzee Lana lexigrams for objects. Later, with his wife Sue Savage-Rumbaugh, the chimpanzees Sherman and Austin were taught lexigrams for all kinds of objects and actions. In the 1980s work by the Rumbaughs started with bonobos, their star subject being the bonobo Kanzi, who eventually learned to use hundreds of lexigrams.
The scientists of these sign language and lexigram studies with nonhuman great apes eventually claimed that their ape subjects had acquired (some form of) human language. Not only had the apes learned that a symbol can refer to something in the outside world, the capacity for reference, but the apes also made meaningful combinations of symbols, even expressing grammar in these combinations, mostly by use of a specific syntactical order of combining symbols. Furthermore, it was claimed that the apes used the symbols for a rich variety of functions, not only requesting objects or food from their human caretakers, but also commenting on the things they saw, and even expressing their emotions and thoughts through their use of symbols. For example, the Fouts claimed that Washoe could use the sign HURT to express that she had a stomach ache. Especially Patterson has claimed that the gorilla Koko expressed her feelings a lot, such as signing HAPPY GOOD YOU COME when visitors arrived, and signing CRY and SAD when her famous kitten All Ball died. The Fouts and Patterson also claimed that the apes used signs such as DIRTY and DEVIL to make insults, for example, when Washoe was denied something by Roger Fouts, she signed DIRTY ROGER.
Meanwhile, another project with a signing chimpanzee was carried out in the 1970s. This was the project by the psychologist Herbert Terrace, who at Columbia University taught signs to the chimpanzee Nim Chimpsky, named after the famous linguist Noam Chomsky. Terrace obtained quite different results than the other researchers and he concluded that the apes had nót learned a human language. The publication of his work caused a lot of upheaval, in what has become known as the ape language controversy.
This controversy drew attention to problems of methodology and interpretation in ape language research. Terrace was the first to carry out systematic video analyses of the sign interactions between the chimpanzee Nim and the humans. To his surprise he found that Nim imitated the signs the humans made to a large extent. Combinations of signs appeared not to show any semantic or syntactic structure, consisting largely of stringing signs together in order to get the human into action. Motivationally, Terrace concluded that signing apes only used their signs acquisitively: in order to acquire objects and actions or attention from the humans, thus using the symbols in a very limited way compared to what human children do when they first start acquiring language, who also ask questions, describe thoughts and express emotions.
What are we to think of this whole controversy, which persists up to this day, with more or less two camps still criticizing each other: those who believe that nonhuman apes have acquired a human language, and those who do not. Who is right?
When I studied psychology I became very intrigued by the field of ape language research. I was less interested in the question whether the apes had acquired an actual language. My interest was more in the possibility that this form of symbolic communication could be a new and third way to obtain new knowledge about the consciousness or inner nature of the apes: their thoughts and emotions. Scientists already analyzed the nervous systems and nonverbal behaviour of animals in their study of animal consciousness, but the ape language studies promised a verbal and direct expression of their inner conscious states. During my psychology studies I visited the Chimpanzee and Human Communication Institute of Central Washington University, where the Fouts continued to study the sign use of Washoe and four other chimpanzees. I also visited the other ape language researchers and returned to Holland, determined to set up a Ph.D. project about the possibly linguistic apes. Working from the philosophy of science section at the Philosophy Department of the University of Nijmegen, I paid multiple visits to the States in the 1990s to collect data about the apes in the language projects.
My first focus was on the use of what in child language research are called internal state terms: words or signs that refer to Positive affect, such as happy, love and funny, Negative effect, such as angry, cry, sad, Physiological states, such as hungry, thirsty and tired, Perception: see, hear, smell, pain, cold and Cognition: know, think, remember. Human children start acquiring internal state terms when they are 2 years old and start using these for other people when they are 2.5 years old. According to the Fouts, the signing chimpanzees had acquired signs that referred to the chimpanzees’ internal states, such as HAPPY, SAD and CRY. However, when I investigated the archives of the signing chimpanzees from 1970 to 1995, I found that almost all of these signs had not been used at all in these 25 years. The signs that they did use could all be explained as not referring to their inner states, but as useful signs to manipulate the humans. The use of CRY, for example, was mostly used by the chimpanzees after they had first made a request for food or an object, which was then denied by the humans, to which they then signed CRY. When I visited Koko and Penny Patterson at the Gorilla Foundation in California, I was given access to all the Koko diaries on microfilm from the beginning of the project to 1995. The discovery I made there, was that Patterson had been unacceptably selective in her presentation and publication of Koko’s sign use. The unedited Koko diaries showed that Koko was drilled in using signs such as SMILE and CRY by showing her pictures of smiling and crying humans many times a day. More importantly, this unedited material also showed the many, many times that Koko used these signs incorrectly. Thus she would sign CRY to a picture of a smiling man and SAD CRY DEVIL to a picture of a horse. Patterson then gave an unrepresentative presentation of Koko’s signing by only publishing her sign utterances that appeared to be linguistic or referring to internal states, while omitting all other clearly nonlinguistic use of signs by Koko. These discoveries made me abandon this focus on possibly internal state signs, because these were either not used by the apes or did not show an actual reference to inner states.
I then changed my study into a large, systematic analysis of all use of signs by Washoe and the other chimpanzees at the Chimpanzee and Human Communication Institute, when they were in interaction with longtime human companions. I hoped to find solid data that could show what the chimpanzees were actually doing with the signs and whether their sign use showed any linguistic properties. I analyzed in great detail 22 hours of videotaped interactions from 1992 to 1999. The subjects of my study were Washoe, who was then between 26 and 34 years old, Moja, who was 19 to 26 years, Tatu, between 16 and 23 years, Dar, 15 and 23 years, and lastly, Loulis, Washoe’s adopted son, who was then between 14 and 21 years old. 3,500 chimpanzee sign utterances were obtained for analysis. I determined the particular individual signs that the chimpanzees used, the kinds of combinations of signs that they made, and each sign utterance was coded for its communicative intention or the reason for which the chimpanzee used the utterance.
These were my results: The individual signs that Washoe and the other chimpanzees used mostly consisted of object or action signs, with the signs DRINK, FLOWER, EAT and GUM among the most used signs. Their combinations of signs showed no semantic or syntactic structure and the longer combinations were quite similar to Nim’s: stringing signs structurelessly together, in order to provoke the humans into giving them something. For example, Washoe made the following 13-sign utterance: DRINK HOT GIMME DRINK TOOTHBRUSH GUM HOT GIMME TOOTHBRUSH GIMME TOOTHBRUSH GUM BOOK. In terms of their communicative intentions, the chimpanzees predominantly used signs when they wanted to request objects or actions from the humans. Thus, Terrace’s characterization of ape sign use as acquisitive, used to acquire things from humans, was very appropriate for Washoe’s and the other chimpanzees’ use of signs. For this reason I gave my dissertation the title GIMME GIMME GIMME. The results that I obtained were remarkable findings. They demonstrated a nonlinguistic use of signs of the same apes of which it had been claimed for decades that they signed linguistically.
A further result that I’d like to mention specifically is the use of signs by Loulis, Washoe’s adopted son. The Fouts had claimed that Loulis had learned over 50 signs from the other chimpanzees (because the humans did not sign in his presence for more than 5 years), which was then interpreted as a cultural transmission of a human language by chimpanzees. In the videotaped interactions that I analyzed Loulis only used the four signs POINT, GIMME, CHASE AND HURRY. These last three signs are normal chimpanzee gestures that have been observed in many chimpanzees in captivity and in the wild. GIMME is the begging gesture with a stretched hand palm up, CHASE the wrist-hit gesture to initiate play and HURRY the impatience gesture of shaking the hands. Thus Loulis had only learned to point to things that he wanted. He never used a sign to refer to objects and did not show a linguistic use of signs.
The different results that I obtained with Washoe and the other chimpanzees, compared to the earlier studies, can be explained by differences in methodology and interpretation. In my study I only used videotaped interactions and I used large corpora or collections of unedited sign use by the chimpanzees. These methods had not been used enough in the previous studies. Many claims from the ape language researchers had been based on unfilmed and unrepresentative instances of signing by the apes, causing unwarranted linguistic conclusions.
Let me explain this in more detail. The problems of using unfilmed data are that these are insufficiently reliable. First, perception errors can occur, mistaking a scratch for a sign or mistaking one sign for another. For example, the signs HURT and BANANA are close to each other. HURT is made by contacting the tips of extended indexes of opposite hands. BANANA is made by the index grasping or rubbing the tip of the other hand’s extended index and pulling it toward oneself. (of ene sign voor andere). A human, focused on finding evidence of linguistic sign use, as all researchers were in these projects, could then mistake Tatu signing TATU BANANA for TATU HURT. Second, memory errors can take place in the time it takes after a signing interaction with an ape to then go back inside and write down the observation. A human might then forget all the exact individual signs that Washoe had made or the correct sequence of the signs. For example, Washoe could have signed BOOK HURRY APPLE ORANGE GIMME RED GIMME YOU WASHOE. An hour or so later the human could remember only something like HURRY YOU GIMME RED APPLE. Third, with unfilmed observations there is no way of assessing the role of imitation of the human by the apes, as there is no reliable information whatsoever on what the humans signed in these interactions.
The fact that large unedited corpora of sign use were used insufficiently means that the human researchers in these projects focused on incidental sign use that appeared to be linguistic, when only by a thorough and systematic analysis of all sign use can one make definitive conclusions. For example, Washoe’s description of a swimming swan as a WATER BIRD has often been presented as an example of a meaningful and creative combination of signs. However, without the analysis of large corpora there is no way of knowing whether Washoe may also have made many unmeaningful combinations of signs such as WATER BANANA and COOKIE BIRD. In which case the WATER BIRD combination would have been just another meaningless combining of signs. When I analyzed the unedited Koko diaries on possible internal state terms this was exactly what I found: Patterson had published only those sign instances of SAD and CRY in which she seemingly referred to her emotions and thoughts, but systematic analysis of all Koko’s sign use showed that she more often used such signs without any reference to her emotions.
Problems of interpretation have also been part and parcel of the ape sign language projects. For example, much has been made by the Gardners and Fouts of the fact that the chimpanzees asked questions, asking for information, using the same accompanying nonverbal behaviours in which humans ask a question in sign language: holding the sign, looking at the other person while waiting for an answer. However, these accompanying behaviours are evidence of intentional communication rather than demonstrating a linguistic use of signs. More importantly, all of the instances in which the apes asked a question can better be interpreted as the apes asking or making a request for food, objects or actions and then waiting for the human to give the ape something. Another example is the interpretation of the sign DIRTY as an insult. DIRTY is made by loudly hitting the chin with a hand, thus making a loud noise by which to attract the attention of a human. DIRTY is then better interpreted as an attention-getter. Chimpanzees in captivity all over the world use many such attention-getting behaviours to draw the attention of their human caretakers.
Besides Terrace’s analyses and my own study, only one further study in ape language research exists that is methodologically and scientifically sound. This is the study published in 1993 by Sue Savage-Rumbaugh of Kanzi. Not of his use of lexigrams, but of his understanding of short English commands, such as “put the carrot in the water” or “put the keys in the refrigerator”. All interactions were filmed and the contents published entirely. The results showed that Kanzi was indeed able to understand hundreds of human words for objects and actions.
So, after almost a century of ape language research what conclusions can we draw today about the capacity for language in nonhuman great apes? Again, in the definition of human language that I use, language consists of 3 important aspects: 1. The use of symbols, 2. The capacity to meaningfully combine these symbols grammatically and syntactically, and 3. A rich variety of functions for which the symbols are used. Using this definition of language I reach the following conclusions: Nonhuman great apes are able to understand hundreds of symbols for all kinds of objects and actions. Something which has also recently been found to be true for border collie dogs, such as the German border collie Rico, who understands over 200 words for toy objects and the American border collie Chaser who was demonstrated to understand even over a thousand words for objects. So the apes can understand symbols. But when the apes combine symbols, however, these combinations don’t show any semantic or grammatical structure. Lastly, the apes almost only use their symbols to acquire things from humans, whereas humans themselves, including human children who start acquiring language, use language for a rich variety of different functions. Thus, I conclude that nonhuman great apes have not been able to acquire language and I claim that this is a qualitative difference in capacities between human and nonhuman great apes. In my next lecture I will discuss what this conclusion means for our ethics towards nonhuman great apes.
Is absence of language in nonhuman great apes morally relevant?
The capacity for language has long been a major dividing point between humans and all other animals. For the 17th century philosopher René Descartes human language was the only certain proof that a being possessed not just a body, but also a mind. Descartes defined language not only as an ability to use words, which he knew parrots could also learn and use. He saw language particularly as the meaningful combination of words or symbols, through which humans can express thoughts and all their inner life. In contrast, in Descartes’ view, birds using words and all other animal communication simply consisted of a mindless expression of the animals’ passions or motivation. Indeed, his assumption that parrots who had learned words only used these to express their passions, was for Descartes a good argument that all nonhuman animals don’t even háve thoughts, a mental life or a mind. Nonhuman animals thus were reduced to mindless automata or machines. The dichotomy between mindful humans and mindless animals thus constructed by Descartes, had a major impact on the way science and philosophy viewed nonhuman animals for centuries. The absence of language in nonhuman animals has since often been used as an excuse for quite speciesist moral philosophies and has led to the justification of practices such as animal experimentation.
Since animal ethics came up in the 1970s as a subdiscipline of ethics, language has continued as a morally relevant ability in several animal ethics theories. I will mention two cases. The philosopher R.G. Frey from Bowling Green State University in Ohio argued that language is a required capacity in order to have beliefs and desires and therefore interests. He expressed this view in his 1980 book Interests and Rights: The case against animals and in his 1983 book Rights, killing and suffering. According to Frey, because nonhuman animals are incapable of having desires and therefore interests, they don’t have rights like humans do and he argued for an inegalitarian view between humans and other animals, where nonhuman animals are of lesser moral value than linguistic humans. The British philosopher Peter Carruthers, now at the University of Maryland, claimed that language is a necessary ability to experience mental states, argueing that nonlinguistic animals do not have sentience or phenomenal consciousness. Sentience was defined by Peter Singer in Animal Liberation as “the capacity to suffer and/or experience enjoyment.” Phenomenal consciousness is a term coined by the British philosopher Ned Block with which he refers to the experiential properties of sensations, feelings and perceptions, thoughts, desires and emotions. Carruthers published his view on animals in his 1992 book The Animals Issue and he has since continued to write further publications about it. In Carruthers’ view nonhuman animals, lacking language and therefore having no sentience, are not part of the moral community. Humans only have indirect obligations to other animals, such as not being cruel to them, because cruelty towards animals shows the presence of a cruel character, which could then also lead to cruelty towards other humans.
In the Great Ape Project (GAP), set up by Peter Singer and Paola Cavalieri, the results from the studies with signing apes were used in the argument for equality among all great apes (human and nonhuman). The project was launched in 1993 with the presentation of the book The Great Ape Project. Equality Beyond Humanity. Several ape sign language researchers contributed to this book with chapters in which they demonstrated the linguistic abilities of signing apes. These were Roger and Deborah Fouts, who worked with Washoe and four other signing chimpanzees, Francine Patterson, who studied the signing gorillas Koko and Michael, and Lyn Miles, who worked with the signing orangutan Chantek. Later, Sue Savage-Rumbaugh, the researcher who studied the use of lexigram symbols by the bonobo Kanzi, joined this call for equality. In claiming that these nonhuman apes had acquired a human language, these researchers argued for an egalitarian treatment of nonhuman apes, similar to how we treat other human apes.
The participation of the ape language researchers in this moral project for equality has had severe negative consequences for one of the signing apes, something which I have been a personal witness to. When the GAP book was published, the signing orangutan Chantek was living at the field station of the Yerkes Regional Primate Research Center of Emory University, near Atlanta. The people in charge at the field station were not amused by Lyn Miles’ contribution to the project and frustrated her attempts to visit Chantek. Fearful of the implications of Chantek as a linguistic ape, Chantek was isolated alone in a cage and neglected. Caretakers were not allowed to sign to him, even though he signed to them about his wants and needs. All of this led to serious boredom in Chantek and in response he ballooned into a very fat orangutan. Linguistic apes were thus seen as a threat by the biomedical establishment, which feared that further experimentation on nonhuman great apes would become prohibited, as it has indeed become in various countries around the world, including the Netherlands, though sadly not (yet) the United States.
During one of the many coffee breaks.
As I talked about in my first presentation this morning, I have analyzed the archives of all the signing ape projects and carried out a systematic study of the use of signs by the famous signing chimpanzees. The results of these investigations were that there was no evidence for a linguistic capacity in these apes. If one considers all the results from the various ape language studies (including those that examine the use of lexigrams by bonobos such as Kanzi), one can conclude that nonhuman great apes are able to understand hundreds of words or signs. But when it comes to combining symbols in a meaningful or grammatical way there is no solid evidence that the ape subjects are able to do so. Also, when it comes to the motivation for (symbolic) communication it seems that there is a qualitative difference between human and nonhuman great apes. The apes in the language studies use symbols more as a tool to request things from humans rather than use it for a rich variety of motivations as humans do. Thus we can see a difference in motivaton for using symbols between human and nonhuman apes. The psychologist Michael Tomasello has suggested that this originates in the fact that nonhuman great apes are more focused on competition in their communication, while human apes are more focused on cooperation. So, until new evidence comes forth, we have to conclude that only human apes have the capacity for language and that this amounts to a qualitative difference between humans and other apes.
Where does this conclusion lead to, when it comes to our morality towards nonhuman great apes? Are nonlinguistic apes due less, morally, than apes who have language?
Let me take a closer look at the positions held by Frey and Carruthers. According to Frey, for a being to have interests, it needs to have desires. But a being can only have desires, if it has beliefs. And for the particular capacity to have beliefs, Frey argued that language is necessary. What are we to make of Frey’s position? Is it true that language is necessary in order for a being to have desires? Let’s specify the subject and focus on a being’s desire not to be in pain. Take a rat in a laboratory who is inflicted pain in an experiment. If I understand Frey correctly, the rat may feel pain, but cannot have a desire to be freed of this pain. However, when we look at the rat’s nonverbal behaviour we can see that she shows signs of being in distress and that she tries to get away from the pain. In another more personal example, a few months ago, we brought our cat to the vet to examine his blood. First he got an injection to make him doze off. However, immediately upon this injection our cat jumped of the vet’s table, ran away and tried to hide in the corner, hizzing aggressively at anyone getting close to him. Attempts to pick him up again were quite fruitless. Does this behaviour not show a desire to avoid pain? Indeed, in animal welfare research a lot of studies have taken place in which scientists look at particular preferences of animals. Famous are Marian Stamp Dawkins’ preference tests of various types of housing for chickens, which showed that chickens prefer soft floors over wire floors, as well as preferring floors with wood-shavings and such in order to be able to dustbathe and perform other natural behaviours. How would Frey explain these preferences? As something else than desires? But does not my cat have a desire to be out of pain and the chickens a desire to live in more accomodating spaces under more natural conditions? Clearly it is possible for a being to have a desire without having the capacity for language. This implies, then, that it is possible for non-linguistic beliefs to exist. Frey disagrees with that assumption by claiming that language is necessary in order to have a belief. This is because in his view you cannot have beliefs, if you can’t distinguish between true and false beliefs, and to be able to do that, you need language. I do not see why, when we look at the past decades of animal cognition research where many nonlinguistic cognitive abilities have been demonstrated in all kinds of animals. This includes the capacity to form beliefs about the physical and social world that surround an animal. In the example of my cat, my cat must have had some form of belief that the vet was only out to hurt him rather than help him, judging us as nasty beings he had to get away from. Frey’s position seems close to the flawed hypothesis of several philosophers of language that language is necessary for thought. This view has long been considered to be incorrect, partly because of all the evidence for cognitive capacities in nonhuman animals, but also because we now ascribe desires and beliefs to nonlinguistic humans such as pre-linguistic infants. In conclusion, then, Frey’s arguments do not hold.
What about Peter Carruther’s view? Frey still ascribed a capacity for sentience to nonhuman animals, while Carruthers claims that all nonlinguistic animals are not even sentient. He argues that a being can only experience something if the being can think and reflect about the various states the being is in. And in order to think and reflect, language is necessary. Again, like Frey, Carruthers links thought or cognition to language. So, again, we can rebut his proposal by showing that nonlinguistic cognition is possible and existing in many nonhuman animals. Indeed, language research with great apes and border collie dogs has shown that though nonhuman apes and dogs do not have language, they are capable of understanding hundreds of symbols and therefore have concepts of all kinds of objects and actions in the world around them.
However, different from Frey, Carruthers claims that language is not only necessary for thought, but also for sentience. He argues that it is the linguistic terms for various conscious states that make them phenomenally aware. Only by having the linguistic labels of pain and pleasure can we experience these states and can we even be aware that these states are different from each other. In Carruthers’ view nonverbal pain behaviour in an animal does not demonstrate sentience. He believes that any behaviour can exist without a mind or a sentient being who experiences these behaviours. He makes reference to being able to drive a car automatically while being absent-minded, and to the medical phenomenon of blindsight, a brain disorder where humans do not experience seeing things, but where further tests show that they nevertheless did perceive things. He even comes up with the purely hypothetical and speculative case of Penelope, a woman who shows abundant pain behaviour, screaming loudly and trying to get away, but without ever feeling or experiencing any pain. I believe that Carruthers has lost himself here in a hypothetical, unreal world. It is true that certain behaviours may occur mindlessly, as in certain reflexes or when we automatically carry out certain actions that we long ago learned. However, Carruthers is taking this fact too far in claiming that even pain behaviour can exist without feeling or experiencing pain. His hypothetical Penelope has never been found by scientists. Instead, all current knowledge shows that human and nonhuman animals react with similar nonverbal behaviours to pain and indeed we share the same pain centers in the brain such as the amygdala. These similarities in brain and behaviour demonstrate that nonhuman animals are just as able to experience pain as humans. Explaining the nonhuman animal behaviour differently risks the peril of not even being able to ascribe pain and other feelings to humans. In conclusion, then, Carruthers is wrong in claiming that language is necessary for sentience. The various phenomenally conscious or sentient states such as pain and pleasure existed before language and only we humans with language then add a linguistic label to these states.
The absence of language in nonhuman animals thus does not imply that they do not think, nor that they do not feel. Consequently, absence of language in nonhuman animals does not call for an ethics in which humans are treated differently from all other animals.
Arguing for equality among all great apes based on the capacity for language, like the Great Ape Project and the ape language researchers have done, however, is also incorrect, now that these new studies have shown that there is no evidence for the linguistic claims of these researchers. A qualitative difference between humans and other apes thus exists, in that only human apes have language.
It is, however, a matter of argument whether cognitive abilities such as language or rich motivations for communication are morally relevant at all or not. When we look at egalitarian morality towards humans, qualitative differences in cognition or motivation are usually considered to be irrelevant for a morally equal status of all humans. In what is known as the argument from marginal cases, humans with mental or physical disabilities are considered worthy of moral consideration just as much as supposedly ‘normal’ humans. This is because in egalitarian views on human morality, the presence of sentience or phenomenal consciousness in all humans is considered to be a sufficient condition for a morally equal status.
In order to avoid the risk of being speciesist by limiting equality to humans only, equality should be extended beyond humans to include all beings capable of sentience or phenomenal consciousness. Thus, equality for all great apes, nonhuman and human, is still arguable, despite the absence of language in our nonhuman cousins. Indeed, current scientific knowledge and opinion shows more and more evidence and arguments for the presence of sentience or phenomenal consciousness in all vertebrate animals and possibly also invertebrates. With sentience or phenomenal consciousness as a sufficient condition, equality should thus be extended to all nonhuman animals.
Before I finish, I would like to say something that is well apt for this conference, where building bridges between science and animal organisations is an important goal. I have found that animal lovers and activists often get very angry and accusative, when scientists claim that there are qualitative differences in cognition or motivation between humans and other animals. Like a red flag to a bull, any qualitative difference between humans and other animals is jumped at with attacks, condemning the messenger scientists as creating a new or confirming an existing moral dichotomy between humans and other animals. I think that this is a very unscientific and above all, very unhelpful way of dealing with the results of animal studies. Humans and other animals have similarities ánd differences and it is senseless to deny this. Just like cheetas are better in running than humans, so can humans be more adept in using symbols than other apes. I hope that the animal liberation and animal rights movements can adopt a truly scientific attitude and reconcile itself with the existing similarities ánd differences among all animals on this planet.
Esteban Rivas's Blog 