20th WCP: Can Deaf Children Be Taught to Think Philosophically?

How is it possible for children to develop a "mature theory of mind"? There are at least three important components in acquiring a theory of mind. First, a child must acquire an ability to use, and understand, certain complex syntactic structures, especially those involved in discourse about mental states (e.g. belief, hope, memory, etc.). Second, a certain level of general cognitive development, in such areas as the ability for reasoning and inference, is necessary. Third, the ability for effective social interaction seems to correlate strongly with the acquisition of a mature theory of mind. (1) The main problem is to explain the intimate and complex interconnections among these three areas of development. Hearing children normally acquire a theory of mind between ages 3-5, and profound changes have been observed to occur around age 4 which have a great impact on the three areas of development just noted.

This problem of how children acquire a mature theory of mind has been the subject of an intense debate among cognitive scientists. One important focus of the debate is the question of priority: Are cognitive/social developments prior to linguistic developments, or is it the other way around? Perhaps, the two "co-evolve" in the course of general cognitive maturation.

According to Leslie, (2) general cognitive maturation is central to the development of a theory of mind. Once a child is able to understand that her beliefs are different from those held by others, she is in a position to interpret others' mental states, like beliefs, motives, and intentions. And this in turn makes social interactions with others possible.

In contrast, many researchers — Astington, (3) de Villiers, Peterson and Siegal, (4) Tager-Flusberg, (5) and Wellman (6) — claim that linguistic developments have a special and central role. According to them, language facilitates and enables the development of theory of mind reasoning. There are two forms this approach can take. The weaker form only claims that a child's proficiency with language about mental states "facilitates" her ability to form mental explanations of behavior. In 1995, however, de Villiers proposed a stronger account, according to which the acquisition of complement structure syntax is an essential precondition for representing propositional attitudes, like beliefs — in particular, the independent truth or falsity of embedded propositions. The ability to discriminate true and false beliefs in others seems a clear case that involves complex syntactic structures embedding clauses with independent truth values. As de Villiers says, "the semantics of complements allow the possibility of explicitly talking about the fundamental distinction between things in the world and things as they are represented in someone's mind". (7)

In this paper, I want to consider some of these issues as they relate to deaf children — in particular, how language acquisition, including the use of mental verbs, influences the development of a theory of mind in deaf children. I will also describe, though quite briefly, how a "philosophy for children" program, such as the one developed by Matthew Lipman, can help deaf children with their mastery of embedded syntactic structures and the use of mental verbs.

I am in sympathy with de Villiers' approach here. I believe that it is fruitful to start our investigation of the development of a theory of mind with a close look at the child's ability to acquire and use syntactic complementation, and see how it provides the child with a new representational capacity for propositional attitudes and mastery of mental verbs. My experience supports de Villiers' approach. In my long years of teaching the deaf, it has become strikingly clear to me that the use of complement structures and mastery of mental verbs not only facilitates but is essential to the acquisition of an ability to formulate mental explanations of others' behaviors. The child may have a hard time learning the relationship between embedded complement structures and the semantics of mental verbs. However, this is essential to understanding the difference between others' beliefs and one's own, and, more generally, without it, it is doubtful that the child will ever develop an adequate theory of mind.

Sentences about mental states generally require an embedded proposition or complement — for example, "I thought that the dragon was fierce ", "Billy hopes he will win the prize" An inner clause is embedded in the whole sentence. It is not like the two independent clauses found in a conjoined sentence, like: "Tico has gold wings, but the other birds don't." Complements have certain syntactic, semantic and lexical properties which the child must learn. Thus in de Villiers' account it is the semantics of these difficult complex syntactic structures that allows for the possibility of discussing the distinctions between reality and our perception of it. It is not until the child can make this distinction that she is able to talk about lying, mistakes or false beliefs.

Thus the development of the child's capability for "pretend play" (8) at about age 2 is considered to be a major steppingstone towards developing a mature theory of mind since this activity helps the very young child to separate subjective mental reality from objective reality. Yet it is not until age 4 that children can actually represent false belief. Why is this ability to represent false belief so crucial to the language and reasoning of the child? Perhaps it is because understanding false belief entails the ability to hold simultaneously two contradictory mental representations in view — one's own belief and the other person's belief which is incongruous with one's belief and with reality. Thus, it is when a child has achieved the ability to understand others' false beliefs (as false beliefs) that she can be said to possess a theory of mind and can use this to comprehend why other people act as they do.

Earlier we stated that a child who is capable of producing the language of complementation, namely, embedded sentences within complex sentences, has a representational capacity for propositional attitudes. This gives the child the ability to discuss the desires, intentions, beliefs, and perceptions of others, and affords her the opportunities to add further mental verbs to her vocabulary, thus enhancing lexical development.

But exactly how does this process take place? What steps are necessary for the child develop a theory of mind along with a mastery of the required complement structures? In a 1994 study, (9) de Villiers listed the following five steps:

1. The child has mastered basic simple sentence forms which correspond one to one with reality. 2. The child must become aware of the discrepancies between some sentences and reality, thus becoming aware of and recognizing pretense as well as mistakes. 3. The child masters the first embedded structures under verbs of communication/mental state/desire: the child acquires the fundamental syntax without regard to the meaning (semantics) within the complement. The child attributes the truth value of the complement as if it were a simple sentence independent of the main verb. 4. The child begins to notice situations where the verbs of communication suggest that the complement is false when embedded. 5. The final step occurs when the child becomes fully proficient with the semantics of embedded sentence structures with mental verbs. In particular, this involves the child's appreciation of the fact that mental verbs do not denote overt actions, and that she must infer others' mental states from observable actions including linguistic utterances. It also involves the child's ability to communicate her thoughts to others.

In developing a theory of mind, there appear to be two distinct yet intertwined themes that when taken together form a whole. First, the child need to be able to represent things, in particular, represent cognitive and metacognitive states (including metalinguistic activities). In order to accomplish this, the child must master the use of sentences embedding complement clauses. A central element of this mastery is the understanding of the relationship between the truth values of embedded sentences and those of the embedding sentences — in particular, an appreciation of the fact that the truth or falsity of an embedded sentence is independent of the truth or falsity of the embedding sentence, and also of the speaker's belief. Obviously, this is essential to the child's representing other people's beliefs and perceptions as well as her own. And it is clear that this ability is a precondition to her successful social interactions with others. It is also clear that what makes this possible is the linguistic resource of syntactic structures with embedded sentence clauses with an appropriate semantics — semantics that makes clear the truth-value relationships between the relevant syntactic units.

The second thread in the development of a theory of mind is "mental holism" — the fact that our mental concepts constitute a seamless and intricately interdependent system. It is incoherent to think that a person has mastered the notion of assertion without mastering the notion of belief: if you assert (say, "It is raining"), then you must have the corresponding belief (that it is raining) — or at least you represent yourself as having it. If you master the concept of pretending you need the concept of belief and make believe. If you are to acquire the concept of promising, you must possess the concepts of believing, intending, being responsible, and a host of other mental concepts. And think of the variegated sets of mental concepts involved in the concepts of, say, envy, embarrassment, pride, and the like. Although it is not necessary to use the word "promise" when you promise, the fact is that without learning a system of mental verbs, it appears just not possible to master this intricately holistic system of mental concepts. With deaf children there is the following suggestive fact about mastery of words and mastery of concepts: even though they are able to assert something, say with gestures, if they are to learn the mental/metalinguistic verb "say" (and acquire the concept of assertion) they need to learn it as they carry out the act of saying. As children begin using mental verbs that express mental concepts — think, pretend, promise, lie, and the like, they engage in mental activities that are essential for social development. Many of these activities, especially those involving interpersonal interaction, seem to require the possession of the corresponding concepts. One can't promise if one doesn't know what it is to promise

Earlier we pointed to the systemic interconnections among mental concepts. We now turn to the matter of how children learn the relationships between certain metalinguistic/metacognitive verb pairs, such as say and think ; ask and want ; deny and doubt ; promise and intend" (10) and how competence in usage of these kinds of verbs is related to the child's becoming literate. Olson and Torrance found that when children gave justification for their response, they appealed to "know" more than either "think" or "pretend", and "think" more often than "pretend". In observing the following pairs of words in children's everyday conversation, imagine/pretend, realize/remember, doubt/deny, wonder/decide, it was clear that children found that "wonder, remember and doubt were among the easiest to define, whereas imagine, decide, pretend, and deny were the most difficult to define". (11) From the research it seems that children's ability to offer justification of choices of definitions for the correctly chosen verb was the best indication of the children's competence with these verbs. According to Olson and Torrance, there is a hierarchy of difficulty in comprehending these pairs of verbs, going from most difficult to least difficult: "doubt/deny: imagine/pretend: realize/remember, then, wonder/decide." (12) They found that the child's response to a mental sentence with an embedded clause is influenced both by the mental verb of the embedding sentence and the plausibility of the complement clause.

An interesting point made by the researchers in this area is that children who have competence in using metacognitive verbs such as "think", "know", "decide", and "remember" in their conversational interactions with others tend to read earlier than other children even though different social groups have varying levels of competence in usage of these verbs.

Successful reading of narrative stories requires the child to possess not only the ability to monitor one's own comprehension, but also the ability to make judgments and inferences about the emotions and intentions of the characters within the story. But in order to be able to do this the child must grapple with mental verbs and their meanings. (13) Many of these words tend to have fuzzy boundaries making clear meanings difficult to define , and in order determine whether a child has properly acquired a mental verb we must look at the relationship between how the child uses the mental verb for herself and her ability to use it in her communication with others. It's no wonder then that comprehension of embedded complement structures following mental verbs is difficult for all children but particularly for the deaf child. Perhaps this is one more complication which compounds the difficulty of the deaf child to understand complementation and to comprehend complex, abstract, narrative texts.

When one considers metalinguistic and metacognitive verbs, the children's performance on the usage of these verbs correlates with other aspects of oral and reading competence. The better the children are able to use these verbs in their everyday conversation, the better readers they are. There has been much speculation regarding the fact that these verbs may have been present in the child's vocabulary since the age of three. As has been claimed, "Basic to this system are the verbs which mark an understanding of the relation between the speaker's meaning as well as verbs which mark the psychological commitment of the speaker to what is said, verbs such as know, think, guess, believe, doubt, imagine and so on. It may be that in learning to cope with a literate environment, the child comes to master and use concepts that permit him or her to deal with both language and thought in a new way, so that language and thought can be differentiated and decontextualized. Language no longer merely expresses the intention of the reader or writer, but comes to have a meaning in its own right, what we have called sentence meaning." (14)

The connection between mental concepts and the embedding of cognitive structures and mental verbs is crucial. The child cannot learn these systematically interconnected and interdependent concepts. Somehow the child must acquire and internalize a whole system of highly interconnected complex concepts. Exactly how that is accomplished is an empirical scientific question, but it surely must be counted among the most impressive of human cognitive achievements. At any rate, it is difficult to see how this is possible without the support of a linguistic structure adequate to represent the intricate patterns that govern our mental concepts.

Let us turn now to see how the deaf fare in the development of mental concepts and the embedding of cognitive structures and mental verbs. The process for hearing children appears to develop smoothly without hitches for the most part, but for the deaf child the process is spread out over a much longer period, with bumps and many obstacles. In the previous section, we were concerned with the development of theory of mind reasoning in hearing children. Now we must consider how or whether this information can be applied to deaf children. We want to know what it's implications are for the deaf children so that we can perhaps help them in developing a mature theory of mind and using it to reason about others' mental states. It would appear that if a child experiences normal language development, then her theory of mind reasoning will develop normally as well. However, since deaf children are significantly delayed in their language development, one can predict that the deaf children will be significantly delayed in theory of mind reasoning as well. It has been shown in a study by Peterson and Siegal that severely and profoundly deaf children between the ages of 8 and 13 were significantly delayed in mastering false belief reasoning. In fact, they concluded that the acquisition of a mature theory of mind is dependent on fluent conversational experience, in particular, communication about thoughts, feelings, and desires. Remember that this entails understanding cognitive concepts, mental verbs and embedded-clause structures. This information is based on early language interaction in families of deaf children whose hearing parents lack fluency in sign language.

The outcome of the study proved that hearing children between 3 and 4 years of age develop the relationship between seeing and knowing, yet it is only at 4 years old that the child learns to discriminate between the knowledge of the seer and the nonseer. In experiments such as changed location, the unexpected object in a familiar container, the sticker-finding game, and explanation of action task, each task ended with a question about false belief. The results of the experiments found that deaf children with a mean age of 7.41 correctly identified the false belief question. The most interesting of the tests was that for spontaneous explanations of actions on videos which were scored with respect to contents having references to cognitive states such as thoughts, beliefs, knowledge, and ignorance or for other mental states like desires and emotions. Since giving a cognitive explanation of action necessitates the use of an embedded complement construction, this language elicitation task was (de Villiers felt, and I would agree) the strongest predictor of the child's theory of mind performance on both the verbal and nonverbal tasks. It was concluded that oral deaf children's successful performance on both verbal false belief tasks, and the non-verbal hide-and-seek task was significantly delayed relative to hearing preschoolers (by approximately 3 years on average). Further it was noted that the reason for this 3 year delay in developing mature theory of mind reasoning was due not just to the language-for-the-task, but rather a delay in a deeper conceptual understanding that underlies both verbal and non-verbal tasks. This would not seem accountable by Leslie's theory which claims that the mature theory of mind is essentially independent of language acquisition. Therefore it was concluded by de Villiers that "mastery of the syntactic process of complementation allows for the "decoupling" of propositions, and hence the representation of false belief." (15) It was predicted that deaf children with early mastery of ASL syntax should have a decided advantage over oral deaf children since they would be exposed to fluent complex language system at an early age, enabling them to develop false belief reasoning and theory of mind reasoning sooner.

Although I agree with their findings, questions arise as to how oral deaf children as well as deaf children who converse in ASL can become fluent and competent in comprehension and usage of mental verbs and syntactic complementation. One should not readily make the claim that just because deaf children who use ASL have a language with its syntax and semantics, that they are necessarily competent users of this language or that they are fluent in comprehension and usage of mental verbs, particularly if they have not been exposed to these structures and mental concepts in a variety of contexts. In addition one must factor in the problems of the deaf child who has acquired ASL language having difficulties being able to read and understand English which has a completely different syntactic structure, thus involving problems with translation with understanding the mental verbs and syntactic complementation.

It is a well known fact that deaf child exhibit a significant delay in language development and in developing a mature theory of mind, and some never attain this. Studies have been done showing that deaf children show very little reference to mind, and no apparent reference to false belief. "Certainly the limited speech, vocabulary, and syntax of profoundly deaf children at age four years is insufficient to support any such talk (about the world), especially reference to other's beliefs. It is not yet clear that the subtleties expressed in mind-talk can be accomplished with a non-conventional gesture system, unless the listener has rich contextual support." (16)

One problem is that the deaf children appear to lack access to express states of mind of themselves and of other persons using mental verbs. This in turn means that deaf children lack the linguistic means to talk about it themselves. However, I believe that not only is it premature, but it is presumptuous, to predict that all deaf children, born to deaf parents, who are exposed to ASL, will necessarily be exposed to conversations containing mental verbs, and mental events making this language fully accessible to them. This depends a great deal on the proficiency of the signing adults with whom these children come into contact. Just what are the levels of proficiency and competence necessary for a child to become fluent in comprehension and usage of this language, and what in fact does the input of ASL look like with respect to structuring talk about mental states and mental state verbs? One can also raise the question that if the level of linguistic competence for the deaf child is not adequate to capture false belief using complementation at age four, we need to question which is of more importance: the ability to experience talk about mental events, or to have the representational means that the syntax of embedded complement structures provides. Or are both equally important?

The implications of this for the deaf child are profoundly important. If the assumption that one must have access to the complete language, and that this is essential for the development of a mature theory of mind is correct, then since deaf children will be delayed in developing a mature theory of mind to the extent of their delay in language acquisition, this in turn will impact on the literacy and comprehension of narrative stories and the child's understanding of the possible differences between mental verbs such as promise, intend, think and know. Even these verbs which were deemed to be difficult to understand for hearing children are even more ambiguous and less perspicuous to the deaf child. One questions whether it is possible for the deaf child to ever develop full command (access) to a given language, within a supportive environment, and who are not handicapped by the level of language competence of the family? In the optimal situation it could possibly happen. In most cases the expressions of desire or mental states rarely appeared in test situations, which leads one to believe that the language system developed by the deaf children was inadequate to give them the basic capability to engage in discourse about mental attitudes like false beliefs that characterize conventional talk occurring with preschoolers.

Some of the difficulties that ensue for deaf children on account of their late developing linguistic distinctions are these: mistaking embedded clauses as if they were independent clauses and not medial, dependent clauses; falsely reporting what was said in cases where the character in a story tells a lie, that is, failing to construe the correct semantic status of the embedded clause (i.e.,they take the beginning of the main clause, drop its verb, and just add the verb of the embedded clause to it). This clearly destroys the correct meaning of the sentence. In addition as previously discussed, deaf children consider it quite legitimate to make referential substitutions within classes governed by mental state verbs. Deaf children also confuse different forms of complements that carry different meanings. That is the children lack the ability to represent particular propositional content of another person's mind. According to the studies, (17) the verb "know" is easier for children than the verb "pretend". "Pretend" is in turn easier than the verb "think". From this evidence we can gather a kind of developmental hierarchy of mental verbs that can help in teaching these mental verbs and their concepts to the deaf child. It appears that hearing children are indeed reluctant to answer "maybe" to "think" questions. However, I have noticed that deaf children would prefer to say "I don't know" to every question rather than even using the word "maybe".

III

If as we have seen, both metalinguistic and metacognitive verbs are necessary for children to develop competence and mastery of language which in turn allows them to become literate learners, can deaf children whose language development is delayed over three years attain some level of competency? In spite of the fact that this paper presents issues that seem to paint a very bleak picture for deaf children ever developing a "mature theory of mind", I do not believe that this is accurate. I believe that it is not impossible, although extremely difficult, for the deaf child to acquire this kind of reasoning.

How then would one even begin to attack such a huge problem? It is my contention that this major problem can begin to be addressed by exposing deaf children to Philosophy for Children (18) as early as possible, definitely in Kindergarten. One of the difficulties that I have encountered is that many teachers of the deaf tend to feel that Philosophy for Children tries to engage deaf children of ages 3 and 4 in concepts, words, and dialogue that are too difficult, and too confusing for children of this age. It was claimed that concepts such as beauty, fairness, death, distinguishing various characteristics of humans, nonhumans, animals and logical reasoning and the process of philosophical inquiry are not really appropriate and should wait until later to be addressed, that is, when the children were older and could better understand what was going on. Waiting until later is simply not acceptable. Time is imperative if we want the children to become competent users of complex language. Philosophy for Children as a program does not give lists of skills which once learned in isolation without a context are later promptly forgotten. Rather this program encourages the child, hearing or deaf, to think, to inquire about friends, nature, things in their environment, relationships that exist or are created, and to acquire and use language and mental verbs, to have the capacity to make judgments, to ask good questions and give reasons, to become an integral part in a community of inquiry in which dialogues are of a philosophical and logical nature allowing students to create generalizations, make judgments and then be able to support one's own perspective along with developing a sense of self-correction.

This program (P4C) is composed of a set of varying level novels including manuals which contain discussion plans and exercises for numerous philosophical concepts, ideas, and mental verbs. It is these discussion and exercise plans that make it possible for the deaf child to start on the road to developing a mature theory of mind. The exercise and discussion plans such as those labeled "Mental Acts", or "Understanding and Knowing", "Stating, suggesting, and inferring", "Thinking with ifs", "What is beautiful", and so on, list sentences involving each of these concepts in a particular situation and asks the children to think about each case and tell whether or not they agree with it and then, give a reason as to why. For example, under the concept of "mental acts", you would find questions like: "Do you believe everything you are told?" and "Can you remember and forget at the same time?" or the child would have to delineate between statements, questions, or commands. By applying specific exercises to certain philosophical concepts, I have found that deaf children can begin to use mental verbs and embedded clauses. But they must be purposefully and constantly exposed to these structures and others in conversations regardless of the mode of conversation whether it be oral or through the air (that is signed ASL). The difficulty is that deaf children must not only be exposed to these concepts, ideas and mental verbs, but, in fact, the various distinctions and subtleties of meanings must be also taught to them

Questions taken from exercises in Lipman's manual, such as: "What should people look like? Is a man a person? Is a man an animal?", allow children to ponder what a person should look like taking into account similarities and differences, and this can be carried over to stories which contain philosophical content, such as: "Everybody knows what dragons look like;" "You are not a dragon!"; Everyone can see that you are only a dusty old wanderer."; and "But, best of all, we know what a dragon looks like. He is a small, fat, bald, old man." (19) After reading the story, discussing interesting ideas, and doing the above mentioned exercises, eight year old Kevin stated: "All people, all dinosaurs, all dragons, all kids that have no water will die forever, but are dragons and dinosaurs real?" After three years of being exposed to and doing Philosophy for Children, Kevin not only uses mental verbs, and embedded syntactic structures appropriately, but he is also comprehending and using logic.

We should never underestimate what deaf children are capable of, they can reach a level of competence in usage of mental verbs, embedded complement structures, and false belief through philosophical inquiry and philosophical dialogue. Deaf children can begin to understand, reason and develop a mature theory of mind.