Davidson, Language, and Thought
by Anthony Peter Iannini | last edited: 06.11.2011 | originally written: 2001 |
|
|
Essay Overview:
_____________________________________________
This is an extensive essay on linguistic meaning and whether or not a being can have thought without also having language in order to structure their mental representations. The issue of the existence of degrees of thought types in various animals such as primates and in different types of minds is examined as well.
This essay concentrates on the philosophical ideas of Donald Davidson with emphasis on his views about meaning. There is also some investigation into a model of meaning and linguistic reference as put forth by the philosopher Paul Grice. For clarity, I have put Davidson's quotes in pale green and other people's quotes in light blue. Section headings are written in bold salmon.
I argue that Davidson's "all or nothing" view of language being required for thought, and his belief that only humans have the ability to think in any way like what we mean by the word "thought", are both too strong of requirements because lesser linquistic abilities in other species very well may yield types of thought that exist to some lesser degree. I conclude that thought is not just a simple concept but rather something best represented across a spectrum of possibility.
0. Introduction: In this essay, I examine the issues involved in answering the following two related questions posited by the philosopher and linguist Donald Davidson, "What is the connection between thought and language?" and ". . .can there be thought without speech?" (1984, p. 155)R. My examination of the issues surrounding these two questions, in light of Davidson's positions, will be divided into four sections.
In section one, I explain Davidson's fundamental concept of radical interpretation in order to provide the necessary theoretical background for his answer to the problem of how language influences thought. For Davidson, radical interpretation is the process through which interpreters of a language come to grasp the meaning of that language by utilizing a theory of meaning. Also in this first section, I summarize Davidson's various views and assumptions concerning interpretation and the conditions for a proper theory of meaning.
In section two, I initially define the terms that are essential in this investigation. Then, I outline a spectrum of views concerning the extent to which language affects, supports, or is equivalent to, thought. I will then place Davidson's core assertion, his belief that to have a thought requires interpretation of a language, on the spectrum of possibilities. Also, I will summarize the central reasons behind Davidson's claim that being an interpreter of language is essential to being a creature with thoughts.
In section three, I critically examine Davidson's position concerning the relationship between language and thought. I will argue that Davidson's two main arguments for the dependence of thought on language are overly restrictive and ultimately fail to account for all types of thought..
In section four, I summarize my views for and against Davidson's strict theoretical position and I draw conclusions about the relationship between language and thought.
1. Davidson's Theory of Radical Interpretation: Central to understanding Davidson's views about communication and the way in which beings come to understand each other is his notion of interpretation. For Davidson, interpretation is the process of applying a theory of interpretation to the linguistic behavior of some being that yields the meaning of such a being's utterances. For example, if an English-speaking adult says to you, "There is a rainbow outside", then there must be some way in which the sounds created by the speaker come to be meaningful to you.
Strengthening the concept of a general theory of interpretation, Davidson argues that we must formulate a theory of radical interpretation in order to capture the nature of communication (1974, p. 309-323)R. Radical interpretation is defined as interpretation in which there is no shared language or convention between two interpreters. How two such interpreters can come to communicate meaningfully should be the descriptive goal of any theory of interpretation according to Davidson.1
1.1 Davidson's Rejection of Alternate Theories of Interpretation: The problem underlying any theory of interpretation, the correct one of which humans seem to utilize so effortlessly, is determining what type of theory of interpretation we must have in order to derive meaning from the linguistic behavior of others.
According to Davidson, "...we need the theory before we can recognize evidence on its behalf" (1974, p. 142)R. In examining the adequacy of various theories, Davidson rejects those theories that attempt to base interpretation of linguistic behavior on either (a) the meaning of the utterance independently of the speaker's intentional states or (b) detailed information about the intentional states of the speaker or interpreter (1974, p. 142-144)R. Intentional states are defined as the internal mental states of a being, such as beliefs and desires, that are about or represent something.
The rejection of (a) as a theoretical basis for interpretation is Davidson's rejection of convention as useful in understanding how meaning is achieved in communication. Basically, (a) appeals to the meaning of words and sentences that can be derived, relative to some linguistic community, from conventional sources such as dictionaries.
He does not deny that convention may be pragmatically important for various communities of beings that share similar languages, but rather that such conventions are not useful in an attempt to formulate a theory of radical interpretation. In order to clarify this point, consider the following lines by Davidson scholar Simon Evnine:
Davidson argues that communication does not require convention. This can be brought out by imagining two people who speak no common language being stranded together on a desert island. The fact that they speak no common language is no principled (as opposed to practical) objection to their communicating... Radical interpretation shows how this can be done without already understanding his language, without sharing conventions about what sentences mean (Evnine 1991, p. 107)R.2
Davidson also rejects (b) as a theoretical basis for interpretation. This is despite the fact that attempts to develop a theory of interpretation based on the intentional states of the speaker (or interpreter) have gained notable acceptance since the work of Paul Grice. Grice's early views of meaning are exemplified in his following formulation of what it is for some person A to mean something by a sentence x:
A must intend to induce by x a belief in an audience, and he must also intend his utterance to be recognized as so intended. But these intentions are not independent; the recognition is intended by A to play its part in inducing the belief, and if it does not do so something will have gone wrong with the fulfilment of A's intentions (Grice 1957, p. 45)R.
It should be noted that although the above passage describes the essential features of a theory of interpretation that is grounded in the intentional states of the speaker, much more sophisticated hybrids and revisions of the above formulation of meaning have been made.
However, regardless of the sophistication of the particular Gricean model in question, Davidson finds all such models based on (b) to be inadequate because there is an unresolvable problem in trying to describe or define detailed information about a speaker's intentional states (beliefs, desires, etc.) without using linguistic terms. Davidson writes that, "There is a principled, and not merely practical, obstacle to verifying the existence of detailed, general and abstract beliefs and intentions, while being unable to tell what a speaker's words mean" (1974, p. 144)R.3
Davidson's rejection of both (a) and (b) as grounds for a theory of radical interpretation leads to the problem of an interpretation circle. On the one hand, we can not base a theory of interpretation of the meaning of some utterance independently of the psychological states of the speaker. And, on the other hand, we can not base a theory of interpretation on the psychological states of the speaker independently of the meaning of the utterance. Davidson writes of this predicament:
Since we cannot hope to interpret linguistic activity without knowing what a speaker believes, and cannot found a theory of what he means on a prior discovery of his beliefs and intentions, I conclude that in interpreting utterances from scratch- in radical interpretation- we must somehow deliver simultaneously a theory of belief and a theory of meaning (1974, p. 144)R.
In order to escape this circle of interpretation, Davidson must utilize some criteria that is neutral to both the meaning of an utterance and the intentional states of the speaker of the utterance.
1.2 The Truth-Conditional Theory of Meaning: Davidson's solution to the interpretation circle is a truth-conditional theory of meaning. A truth-conditional theory of meanining holds that an utterance x means the conditions or state of affairs that, if realized, would make x true. Such a method of interpretation conveys that an utterance of the sentence "It is raining" by some speaker x at time t is true if and only if it is raining near x at time t, for speakers of English.
Davidson's truth-conditional theory of meaning is formulated with respect to Alfred Tarski's explicit syntactical definition of truth in formal semantics (Tarksi 1944, p. 341-376)R. Tarski's Convention T requires that a theory of truth, which utilizes a predicate such as 'is true', be syntactically formulated so that all sentences of a certain form follow from it. For example, any sentence of the form "x is y" is true if and only if x is y would be sufficient.
However, Davidson is interested in a theory of truth rather than just a formal definition of truth like the one put forth by Tarski. Just knowing the syntactic form of a true sentence does not help in understanding why the sentence is true, because in order to derive information from the truth of a sentence we must already know the object language in which the sentence is uttered. But, this would violate the necessary conditions for radical interpretation.
Therefore, says Davidson, "...in order to accommodate the presence of demonstrative elements in natural language it is necessary to relativize the theory of truth to times and speakers (and possibly some other things)" (1974, p. 150)R. The test of the truth-conditional theory of meaning will be the verifiable environment in which a certain utterance is held true, or, in other words, the extensions of the words in a sentence.
He writes of this, ". . .the evidential base for the theory will consist of facts about the circumstances under which speakers hold sentences of their language to be true. Such evidence, I have urged, is neutral as between meaning and belief and assumes neither" (1974, p. 152)R.
Davidson requires that two additional principles be assumed in order for his theory of meaning work. Firstly, Davidson must assume that agents generally utter things they take to be true (the principle of charity) and secondly, agents generally act to maximize the overall utility of their actions (the principle of rationality).
The principle of charity is necessary because it serves as a basis on which to accept the utterances of a speaker as true relative to the time and the environmental circumstances of the utterance. Davidson writes that, "I propose that we take the fact that speakers of a language hold a sentence to be true (under observed circumstances) as prima-facie evidence that the sentence is true under those circumstances" (1974, p. 152)R.
The principle of rationality is necessary to understand that "...the interpretation of speech must go hand in hand with the interpretation of action generally, and so with the attribution of desires and beliefs" (Davidson 1974, p. 154)R.
At this point, there has been no explicit mention of Davidson's position concerning the relationship between language and thought. However, his theory of radical interpretation and his theory of how truth develops are both critical in understanding Davidson's position about thought requiring language.
2. Language, Thought, and Davidson: Because the concept of language is central to the following analysis, I will briefly define the word 'language' as it is used in Davidson's (and others) discussions of its relationship to thought. |
|
"Luminous Neon Moths", 2011, by Anthony Peter Iannini
|
2.1 A Definition of the Term 'language' as it is Used in the Debate: A language, most generally, is any set of symbols. Symbols are considered parts of the world [see PoW] that mean something to the interpreting system when perceived by such a system. What the term 'language' refers to within the language and thought debate, however, is a natural language. A natural language is a specific type of language, such as Swahili or American Sign-Language, that is used by a community of beings to communicate.
A natural language can be defined as a set of symbols that:
(i) are used for communication amongst beings in some community of beings C.
(ii) are produced by members of C according to a syntax that allows for the recombination and relation of symbols to form more complex symbols.
(iii) include symbols that refer to (at least some) syntactical processes.
Condition (i) is the condition that the symbols in a natural language be used by a community of beings for communication with other beings in the community. For communication amongst members of a community to be successful, there must be some similarity between the interpreted meaning of symbols by each individual member of the community. And, any community of beings that uses symbols for communication must be able to produce such symbols. Symbols could be spoken words, gestured actions, emitted smells, or any other state of the world that can be produced by a being for the purpose of communication with another member of the community to which the being belongs.
If we look to the animal kingdom for communities of creatures that satisfy condition (i), we find a number of salient examples. For instance, some vervet monkey communities have been observed by ethologists engaging in communication through the production of symbols that stand for states of the world. Consider the following passage about the abilities of some communities of this primate species:
...they [vervet monkeys] are able to warn each other of the approach of various predators by distinct alarm calls. If leopards (or other mamallian carnivora) approach, they give the so-called leopard-alarm. Upon hearing it, the other vervets flee to nearby trees where they are safe from leopards. If, on the other hand, vervets perceive the so-called eagle-alarm, they look up to the sky and escape, if necessary, into leafy bushes...two further warnings were classified... as snake-alarm and baboon-alarm calls... (Seyfarth, Cheney, Marler 1980 p. 1070-1094)R.
Ethologists have identified four distinct symbols (verbal calls) that mean four distinct predator threats in the case of vervet monkeys. But, if we continue with the criteria for a natural language as opposed to the language of the vervet monkeys, we can see the deep gap between their use of symbols and the use of symbols by normal adult humans.
Condition (ii) for a natural language is the condition that the individuals in a natural language community must be capable of recombining and relating symbols according to rule-like processes in order to form more complex symbols. These rule-like processes for forming more complex symbols have been referred to as the syntax or grammar of a natural language, and are essential for distinguishing the natural language of humans from the communicative linguistic capacities of many other social animals such as vervet monkeys.
Condition (iii) is that there must be symbols for these syntactical processes of recombination and relation of symbols in order to form more complex symbols. English symbols for recombinatorial syntactical processes include words such as 'and' and 'or' while symbols for relational syntactical processes include words such as 'in' and 'on'. Notice that these symbols do not refer to anything external to the creature using such terms. Rather, these terms are special in that they refer to internal processes that the creature possesses in order to manipulate, arrange, and combine symbols which stand for internal and external things.
Therefore, in order to process symbols according to a syntax, there must be some cognitive ability that allows for the processing of symbols in ways referred to by the combinatorial and relational terms. The answer to whether or not this cognitive syntactical processing ability is prior to or dependent upon natural language will play an important role in determining an adequate response in the debate over the relationship between language and thought.
2.2 The Spectrum of Positions on the Language and Thought Issue: Before considering Davidson's position concerning the nature of thoughts and the requisite conditions for having a thought, I will briefly present the spectrum of views that exist pertaining to the relationship between language and thought. I do this in order to place Davidson at one end of this spectrum and in order to elucidate Davidson's understanding of what it is to be a thought against the background of other positions.
Answers to the question of how language is related to thought fall into two major categories as shown in fig.1 above. For contrast, I begin by presenting the polar ends of the two main categories.
The first category is the communicative conception of language. The strongest communicative positions about language hold that all thought is independent of language and that language is only a medium in which thoughts can be expressed for communication.
Under the strong communicative view, thoughts must have a non-linguistic semantics which linguistic semantics are used to convey.4 The second category is the cognitive conception of language. The strongest cognitivist positions about language hold that language is necessary for all thought and that language is the medium in which all thoughts occur. Under the strong cognitivist view, no animals or infants possess any thoughts (Carruthers and Boucher 1998, p. 1-12)R.5
2.3 Davidson's Position in the Language and Thought Debate: In what follows, I place Davidson's views about the relationship between language and thought within this spectrum. A summary of Davidson's position concerning the issue is summarized in his following statement:
We have the idea of belief only from the role of belief in the interpretation of language, for as a private attitude it is not intelligible except as an adjustment to the public norm provided by language. It follows that a creature must be a member of a speech community if it is to have the concept of belief. And given the dependence of other attitudes on belief, we can say more generally that only a creature that can interpret speech can have the concept of a thought.
Can a creature have a belief if it does not have the concept of belief? It seems to me it cannot, and for this reason. Someone cannot have a belief unless he understands the possibility of being mistaken, and this requires grasping the contrast between truth and error- true belief and false belief. But this contrast, I have argued, can emerge only in the context of interpretation (1984, p. 170)R.
The above passage conveys the essential features of Davidon's argument for the fundamental dependence of thought on language. In terms of the spectrum of views depicted in fig.1, Davidson is at the strongest end of the cognitivist conception of language. His position concerning language's influence of thought can be described as both universal and conceptual.
Being a cognitive universalist with respect to language's role in thought is equivalent to the view that all thought is dependent upon language in some way. Also, being a conceptualist in this context means that he believes it is conceptually and logically necessary that all thought be dependent, in some way, upon language. In order to examine Davidson's position more closely, I outline his argument as follows:
(a) the concepts of truth and falsity can exist only in a being that is an interpreter of language
(b) the concept of a belief can exist only in a being that has the concepts of truth and falsity
(c) in order to possess a belief, a being must possess the concept of belief
(d) a belief is an exemplary thought
(e) therefore, beings that do not interpret language do not have thoughts
I have no quarrel with premises (a) and (b) as very plausible accounts of how the concept of truth and falsity arise. Also, I have no difficulty with premise (d), which is supported by Davidson's view that it is acceptable to limit the discussion to a belief in accordance with his view that, "...belief is central to all kinds of thought" (1984, p. 156)R.6
However, I find that premise (c) is very questionable as a case can be made that (c) is too strong a precondition for a being to have a belief. Because Davidson's overall position in the debate is highly dependent on (c), the following third section of this essay is devoted to an analysis of premise (c).
3. Davidson, Belief, and Concepts: In this section I examine Davidson's premise (c) that in order to possess a belief a being must also possess the concept of belief.
Davidson argues premise (c) in two ways. Firstly, he argues that all beliefs must be at least capable of being recognized as false for their qualification as a belief. I call this his argument from potential recognition.
Secondly, Davidson argues that all beliefs require the concept of the belief or, in other words, a belief requires that the being with the belief also have beliefs about the constituent parts of the belief. I call this his argument from mental holism.
3.1 The Argument from Potential Recognition: At first sight, we may take Davidson's views as the argument that the concept of belief (or the recognition that a belief may be true or false) must be applied to individual beliefs before they can be said to truly be beliefs. In other words, we may think that my belief that "there are nine planets" is a belief only when I have the belief that "my belief that there are nine planets could be false". However, this would cause most of our beliefs to fail the criteria for being a belief.
Davidson can not hold this strict view of each belief by itself without dismissing the vast majority of our beliefs as non-beliefs. In order to combat this problem, Davidson only requires that the being in question be able to, in principle, apply the concept of belief to any particular belief (Davidson 1985, p. 479)R.
For Davidson, a being must be potentially capable of applying the concept of a belief to any individual belief for the being to have a belief at all. For example, I may have the belief that "There is ten dollars in my pocket" without ever having the thought "My belief that there is ten dollars in my pocket may be false".
Why does a creature need to recognize that its belief could be or was false in order to have the belief at all? In other words, why are second-order beliefs about beliefs necessary for first-order beliefs? Davidson writes, "Someone cannot have a belief unless he understands the possibility of being mistaken" (1984, p. 170)R. Also, Davidson has written that to have a belief entails the possibility of surprise, where surprise requires:
...that I be aware of a contrast between what I did believe and what I come to believe. Such awareness, however, is a belief about a belief: If I am surprised, then among other things I come to believe that my original belief was false...one cannot have a general stock of beliefs of the sort necessary for having any beliefs at all without being subject to surprises that involve beliefs about the correctness of one's own beliefs (Davidson 1985, p. 479)R.
We know that a belief is a mental state about the world that can be either true or false and that the being that has some belief that p may be either correct or incorrect. However, the primary question is why does the being that has some belief that p need to be able to recognize that, or be 'surprised' by, its belief that p is a belief that p? I find that Davidson presents no compelling reason for this and that he must acknowledge the existence of beliefs for the concept of belief to be applied to.
I find that there is nothing incoherent in the notion of a being that takes all of its beliefs to be true because it does not have the concept of belief. Such a being could have a belief that p for some time and then come to have the belief that q, where p and q are incompatible beliefs, but also where the being in question never recognizes that its initial belief that p was false. Rather, there is just a transition from the belief that p to the belief that q with no recognition. This line has been taken by Hans-Johann Glock, who writes:
One should grant that a creature can believe that p only if it is also capable of believing something that is incompatible with that belief, notably that not-p. But Davidson has no argument to rule out the possibility of a simply switching from a belief that p to a belief that q, without that switch involving a believing that its original belief was false (Glock 2000, p. 35)R.
These objections to Davidson's argument from the potential of recognition do not, however, apply to his argument for mental holism. I examine this argument for the necessity of the concept of belief for belief next.
3.2 The Argument from Mental Holism: Davidson's argument from mental holism concerning the nature of belief (which is representative of thoughts in general) asserts that, in order to have a belief, a being must be have beliefs about the individual constituent parts of the belief. A collection of beliefs about something is the concept of something. There are two questions that need answering in order to determine whether or not Davidson's argument from mental holism is sound.
The first question I will examine is whether or not non-linguistic beings7 possess concepts. If it turns out that some non-linguistic beings can possess some form of concepts, then Davidson's argument will be weakened significantly. The second question that I will examine is whether or not concepts (or beliefs about the contents of a belief) are indeed required in order to have a belief. If this turns out to be plausible, then Davidson's argument will also be defeated.
3.21 The Question of Non-linguistic Concepts: In order to understand Davison's argument from mental holism as a support for his view that language is required for beliefs (and thoughts in general), it is important to get clear about what Davidson thinks a concept is. He writes that in order to, for example, have the concept of a cat, you:
...must have the concept of an animal, or at least of a continuing physical object... something that can move freely in its environment, something that has sensations. There is no fixed list of things you have to know about, or associate with, being a cat; but unless you have a lot of beliefs about what a cat is, you don't have the concept of a cat (Davidson 1999, p. 8)R.
If we apply this description of the concept of a cat to the concept of a belief, the result is that we can not have the concept of belief without having an array of beliefs about the nature of beliefs. Central to these beliefs about beliefs would be the belief that beliefs can be either true beliefs or false beliefs, which requires application of the concept of truth and falsity- which arises only in the context of being a creature that interprets the language of others.
The result of this application of Davidson's understanding of a concept yields the result that to have either a belief or a concept at all, one must have the concept of belief.
This view of how we come to have any beliefs or concepts at all has been called the "Münchhausen-theory of beliefs and concepts" by Achim Stephan because, as Stephan writes, "As the famous Baron Münchhausen pulled himself out of the swamp by his own forelock, here all concepts and beliefs get drawn out of the realm of the pre-mental by a second-order concept, namely the concept of belief" (Stephan 1999, p. 81)R.
The above description of Davidson's position is exemplary of his philosophical position of holism concerning the mind. A belief only exists as part of a network of beliefs in the being that possess such beliefs. And, this becomes one of his main reasons for denying that we can attribute, for example, a dog with any given belief. Davidson writes of this position that:
We identify thoughts...only as they can be located within a dense network of related beliefs. If we really can intelligibly ascribe single beliefs to a dog, we must be able to imagine how we would decide whether the dog has many other beliefs of the kind necessary for making sense of the first. It seems to me that no matter where we start, we very soon some to beliefs such that we have no idea at all how to tell whether a dog has them, and yet such that, without them, our confident first attribution looks shaky (Davidson 1985, p. 475)R.
In the above passage, Davidson writes, "we have no idea at all how to tell whether a dog has them [beliefs]". In other words, we have no way of verifying the existence of the beliefs, or more generally, the thoughts of a creature, based solely on non-linguistic behavior. Some critics, such as Carruthers, want to point out that there is a deep problem in the kind of assertions that Davidson is making.
Carruthers writes, "…epistemology is one thing, metaphysics is another…That we lack sufficient evidence to ascribe any particular fine-grained beliefs and desires in the absence of linguistic behavior is one thing, whether thought itself can exist in the absence of language is quite another" (Carruthers and Bouncher 1998, p. 24)R.
However, I do not think that objections based on the verificationist nature of Davidson's position are very strong. Rather, I put forth that objections to Davidson's view that beliefs can only exist "...within a dense network of related beliefs" must be attacked more directly (1985, p. 475)R. To illustrate Davidson's holistic view of the mental and why we are not correct in attributing beliefs to animals, imagine that a dog chases a cat into the woods and begins barking up a tree that the dog saw the cat run up.
After a moment or two, the dog continues to bark up the same tree even though the cat is no longer in this tree. We could imagine that the cat has jumped without the dog seeing the cat jump, to some other tree. None of the aforementioned story would be objectionable in Davidson's view because there was no reference to the dog's 'belief' that the cat is in the tree that the dog continues to bark up. However, the question is whether or not we can say that the dog does, in fact, have a belief about the cat being in the tree.
3.22 The Possibility of Evidence for Concepts in Non-linguistic Beings: Davidson would argue, from his position of holism, that the dog does not have a belief because the dog can not be attributed with beliefs that are necessary for the dog to have any single belief. For example, if we say that "the dog believes that the cat is in the tree" then, with respect to Davidson's position, we would also have to attribute the dog with beliefs about the 'cat', the 'tree' and this relationship.
And, this would require that the dog has the concept of a cat and the concept of a tree. How is it that we tell whether or not a being has the concept of something? Usually, the ability to distinguish between things is a mark of conceptualization of the things in question. However, Davidson writes that:
A creature does not have the concept of a cat merely because it can discriminate cats from other things in its environment. For all I know, mice are very good at telling cats apart from tress, lions, and snakes. But being able to discriminate cats is not the same as having the concept of a cat (1999, p. 8)R.
The above passage of Davidson's raises the question of whether or not we are ever justified, on the basis on non-linguistic behavior, in attributing some being with possession of concepts. Davidson also writes that "To have a concept is to classify objects or properties or events or situations" (1997, p. 24)R. According to Hans-Johann Glock, the difference between discrimination and the possession of a concept can be brought out by a Wittgensteininan8 distinction that Glock describes in the following passage:
To be capable of classifying or misclassifying things, a creature a must not just have a disposition to behave in accordance with a rule- as butterflies do when they land only on red petals- but of following a rule. That is to say, the principle which distinguishes Fs from non-Fs must be part of a's reason for differentiating between Fs and non-Fs, not just a law to which its discriminating behavior conforms (Glock 2000, p. 46)R.
The question is whether or not the behavior of any known non-linguistic being is sufficient for attributing reasons for differentiating between something and some other thing in its environment. Certainly there must be some types of non-linguistic behavior that would warrant the attribution of concepts.
It seems obvious to me that we could attribute concepts to an alien being that displays very complex behavior towards the environment and I will not go into a lengthy defense of this point. But what about more mundane creatures here on Earth?
If we recall the example of vervet monkeys presented in section two, we may have reason to attribute rudimentary concepts to such beings based on their behavior. In various ethological and behavioral studies, vervet monkeys have been attributed with behaviors that are indicative of the use of concepts- albeit simple concepts. Young vervet monkeys respond with an eagle-alarm call to harmless birds and even falling leaves before they are capable of making the correct alarm call.
This process of distinguishing between predators in the air and non-predators in the air is reinforced by the older members of the vervet monkey group (Allen and Bekoff 1997, p. 121)R. Also, ethologists have observed that certain vervet monkeys (or vervet monkey groups) have used alarm calls to gain the advantage over other vervet monkeys (or vervet monkey groups) (Seyfarth et. al. 1980, p. 1070-1094)R.
More often than not, the monkeys will check the general area to which the specific predator alarm refers. For example, the eagle-alarm caused a high percentage of the monkeys to look generally upwards to confirm predator calls (Seyfarth et. al. p. 1084-1085)R.
Does the discriminating behavior in the case of certain vervet monkeys qualify as concept formation and execution? The monkeys do seem to have some mode of reacting to an alarm and then checking as to whether or not the alarm corresponds to the circumstances or not. And, they must develop the discrimination skills used to pick out dangerous things from non-dangerous things in the environment.
These seem to be minimally adequate for the possession of simple concepts and simple beliefs. According to Achim Stpehan we can rightly attribute the vervets with beliefs with a concession that, "...it is unclear with which description any belief or want could be ascribed to a vervet monkey..." (1999, p. 89-90)R. Other animal studies, such as those with higher primates being raised in captivity, have shown even further evidence of the sort already discussed9.
"The Vortron", 2011, by Anthony Peter Iannini
|
|
However, my goal in this section has only been to show that there is behavioral evidence, under plausible theories, that strongly suggests some non-linguistic creatures do possess simple concepts on the basis of the aforementioned criteria. Such concepts allow for non-linguistic creatures to have thoughts in the form of first-order beliefs and may even allow for rudimentary beliefs about beliefs.
If either of these possibilities is true, then it will turn out that beliefs and thought, albeit in simpler forms, are possible and independent of a natural language. However, even if a being can not be attributed with concepts and beliefs about the contents of their belief, there is still room for correctly attributing such a being with beliefs. Precisely how this is possible will be considered in the following section.
3.3 The Problem of Attributing Attitudes to Non-linguistic Beings: That we can not know which specific beliefs to ascribe to non-linguistic beings is correct and that Davidson would agree with this concession. However, Davidson would take the problem further and say that we should not ascribe beliefs to the vervets because of the unresolvable empirical difficulty with knowing which linguistically formulated belief we should attribute to the monkeys. |
As Chater and Heyes have pointed out, we can not empirically determine whether or not one of the vervets' calls means 'beware of leopards', 'beware of mammalian carnivores', or 'beware of 98 leopards that find vervet monkeys tasty' (1994, p. 209-246)R. This is the problem of intensionality (with an 's').
The problem of intensionality is that, in normal language, we can have thought attributions that create what are called intensional contexts. For example, we could attribute the following form of a belief to an individual vervet monkey: "the monkey believes that the x is in the sky" where x could stand for 'eagle', 'predatory mammal', or 'dangerous bird'. However, we have no way of knowing which specific description of the thing in the sky would be an appropriate description of the thing in the sky, given the non-linguistic state of the monkey.
With respect to this problem, Davidson writes that "The dog, we say, knows that its master is home. But does it know that Mr. Smith (who is his master), or that the president of the bank (who is that same master), is home? We have no idea how to settle, or make sense of, these questions" (Davidson 1984, p. 163)R.
This is an important point about the problem of attributing any given intentional state, such as a belief, to a non-linguistic creature. As Davidson says, it is very difficult, when speech behavior is not present, to know how to "...distinguish universal thoughts from conjunctions of thoughts, or how to attribute conditional thoughts, or thoughts with, so to speak, mixed quantification ('He hopes that everyone is loved by someone') (1984, p. 164)R.
3.4 The Possibility of Non-Intensional, Non-Holistic Beliefs: It is obvious that, if we are willing to attribute beliefs to a non-linguistic being, there will be some form of disparity between the language we use to describe the belief of the being and the actual content of the belief of the being. However, this problem of accurately describing the content of non-linguistic being's beliefs need not result in the non-attribution of beliefs to such creatures.
A suggestion by Stephan is that we, "...at the very least can attribute intensionless propositional attitudes to vervet monkeys, or perhaps propositional attitudes the intensionality of which is currently undefinable" (1999, p. 90)R. We could construct an attribution such as "The dog believes that the cat went up the tree" and recognize that (i) the dog does possess a belief, (ii) our description of the dog's belief is not equivalent to the content of the dog's belief, and (iii) our description of the dog's belief is an adequate description of the dog's belief based on behavioral information. Davidson seems to reject the plausibility of (i)-(iii) in positing the following:
But what kind of description would suit the dog? For example, can the dog believe of an object that it is a tree? This would seem impossible unless we suppose the dog has many general beliefs about trees... without many general beliefs there would be no reason to identify a belief as a belief about a tree, much less an oak tree (1985, p. 475)R.
Davidson makes the case that we should not attribute the dog with a belief about the tree, or the cat. But, this does not mean that the dog can not have a belief about the whole situation that does not involve the dog having specific beliefs about what we would take to be the constituent parts of the belief. Davidson seems to be overlooking the possibility of holodoxastic beliefs, which are beliefs that are transparent in the sense that they are about the immediate environment and are about themselves with no other 'parts' to have beliefs about.
This position is elaborated by Glock, who writes that "Because of its reliance on behavioral reactions, the holodoxastic move is confined to simple beliefs, notably about perceptible features of the subject's environment. But it suffices to blunt the force of the line 'No thoughts without concepts!'" (2000, p. 43)R. In this way, we can plausibly attribute non-intensional (or intensionally undefinable) and non-conceptual intentional states which could be mistaken and which some non-linguistic being could possess.
4. Conclusions About the Nature of Language and Thought: In the previous sections, I have examined Davidson's main reasons for his position that thought is wholly dependent on language. I have concluded that (1) the concept of belief is not necessary for belief, (2) concepts are not necessary for belief (even though some non-linguistic beings may have concepts), and (3) beliefs can be holodoxastic in nature (non-intensional, non-conceptual, non-holistic). The resultant state which can be minimally called a belief, then, under my interpretation, is any intentional state that can be mistaken.
Admittedly, this description of a minimal belief is very different from the potential beliefs of normal adult humans that are interpreters and utilizers of a natural language. However, I think that even the most advanced language-users, such as ourselves, can have holodoxastic beliefs in certain situations where we are forced into a more primitive state of belief about the environment. Imagine, for example, that you are crossing the street and, unknown to you, a car is speeding towards you.
The oncoming car honks its horn and you jump out of the way. At some point, you must have formed a very simple belief about the car and the danger associated with its speeding towards you. However, it seems plausible that your later description of your belief about the car will not be representative of the content of your belief about the car in the second or so that you had in which to act on your belief.
It would have been wrong to say of you, at the time when you were still in the street, that "you believed that there was a car coming towards you at a very fast pace and that you were in its path and this would cause you bodily injury". If this was your belief at the time, it seems as though you would most likely be dead!
Although this example points to the possibility of holodoxastic beliefs, many of our beliefs, and thoughts in general, are obviously made richer and more complex through the use and interpretation of natural language. Although I think that Davidson is overly restrictive in terms of his conditions on thought, he raises a number of insightful points that are worth mentioning.
Use and interpretation of a natural language, it seems, are required for the concept of truth to develop, for self-representations to develop, and, most likely, for conscious thought and manipulation of thought. With a natural language, we can think beyond our environment, voluntarily, in a way not dictated by the various stimuli of our surroundings. We can form, with language, complex thoughts that are limited only by the constraints of our memory.
But, my point has been to show that thoughts, in a primitive sense, are had by non-linguistic creatures. We can see the complexity of thought as it is gradually influenced by possession of language. This overcomes the serious problem with Davidson's view that thoughts suddenly spring into existence when there is "a dense network of related beliefs" in the relevant being (1985, p. 475)R.
How many beliefs, exactly, constitute a 'dense network' of beliefs? Is a dense network of beliefs a dense network of three, one-hundred, fifty-thousand, or one million beliefs? The proper answer to this, in the context of my position, is that a single belief can be had apart from any others.
Without a language at all, beings can only have holodoxastic beliefs about the immediate environment. The use and interpretation of a proto-language, like that of the vervet monkeys, allows for the potential of some beliefs about beliefs. The use and interpretation of a more complex proto-language, like that of higher primates raised in a human environment, plausibly have more beliefs and can form more complex beliefs that incorporate concepts. As shown in fig.2 above, thought plausibly happens in a spectrum of beings rather than in only one particular type of being.
Adult humans, at the end of the spectrum, can form the most complex beliefs that allow for all manner of novel and complex representations. We could imagine that some creatures, with a highly developed language and memory, would be able to have such a 'dense network' of beliefs that it would make us look like monkeys by comparison. I speculate that this entire spectrum, it appears, is analogous to the stages of child development.
Where does this conception of language fit into the spectrum of positions presented in section 2.2 and outline in fig.1? What I have argued for is at most a weak communicative conception of language and at least a weak cognitive conception of language. This is indicative of my view that both ends of the language and thought spectrum of positions are leaving out important details of the situation.
I find that the possession of thought is much more complex and difficult to attribute than the communicativists hold and that the rudiments of at least some thoughts are independent of language, which is in opposition to many of the cognitivists' positions.
A great number of important questions that I have left open need addressing but are beyond the scope of this paper. One of these questions is, to what degree does the influence of learning a language influence the ability of a human to have thoughts? Another question is whether or not the basic thoughts that I've allowed independently of language are the primary type of thought which language arranges, manipulates, and combines to form higher-order representations.
References:
______________________
Baker, G.P. and Hacker, P.M.S. 1985. Rules, Grammar, Necessity. Oxford: Blackwell.
Carruthers, P. and Boucher, J. 1998. Language and Thought. Cambridge: Cambridge UP.
Chater, N. and Heyes, C. 1994. "Animal Concepts: Content and Discontent," in Mind and Language, 9: 209-246.
Davidson, D. 1974. "Belief and the Basis of Meaning," in Synthese 27: 309-323.
Davidson, D. 1984. Inquiries into Truth and Interpretation. Oxford: Oxford UP.
Davidson, D. 1985. "Rational Animals," in E. LePore and B. McLaughlin (eds.), Actions and Events. Oxford: Blackwell, p. 473-480.
Davidson, D. 1997. "Seeing Through Language," in J. Preston (ed.), Thought and Language. Cambridge: Cambridge UP.
Davidson, D. 1999. "Machines, Plants and Animals: The Origins of Agency', in Erktennis, 51: 19-31.
Evnine, S. 1991. Donald Davidson. Stanford: Stanford UP.
Glock, H.J. 2000. "Animal Thoughts and Concepts," in Synthese, 123: 35-64.
Grice, H. P. 1957. 'Meaning,' reprinted in P.F. Strawson (ed.) Philosophical Logic, pp. 39-48. Oxford: Oxford UP, 1967.
Seyfarth R.M., Cheney D., and Marler, P. 1980. "Vervet Monkey Alarm Calls: Semantic Communication in a Free-Ranging Primate," in Animal Behavior, 28: p. 1070-1094.
Stephan, A. 1999. "Are Animals Capable of Concepts?" in Erktennis, 51: 79-92.
Tarski, A. 1944. "The Semantic Conception of Truth and the Foundations of Semantics," in Philosophy and Phenomenological Research, 4: 341-376.
Footnotes:
______________________
