Why does J. A. Fodor think that there must be a language of thought Is he right - Essay Example

? Why does J. A. Fodor think that there must be a language of thought? Is he right? Introduction Jerry Fodor is a prominent linguistic philosopher and cognitive scientist. Fodor’s linguistic philosophy eschews structural models of language for an approach that builds on Noam Chomsky’s generative grammar. Specifically, Fodor advanced a notion of a language of thought. Fodor’s Language of Thought Hypothesis (LOTH) states that the thought process occurring in the mind is a symbolic system that parallels a language in structure. Jerry Fodor’s presented LOTH in his aptly titled book The Language of Thought (Fodor 1975). Through an examination the major platforms of the LOTH hypothesis, as well as prominent criticisms, this essay examines the extent that Fodor is successful in articulating the existence of a language of thought. Language of Thought Philosophers have formulated LOTH by considering the multiplicity of thoughts, or propositional attitudes. One can identify propositional attitudes in sentences like ‘A wishes that B’, or ‘A thinks that B’, or ‘A intends that B’. A is the subject of attitude, B is the sentence, and ‘that B’ is the proposition or the object of the attitude. If we say that C stands for verbs like ‘wishes’, ‘thinks’, and ‘intends’, then the propositional attitude sentences would take the form: ‘A Cs that B’. Therefore, LOTH can be considered as a hypothesis that features how we create propositional attitudes in our ‘mental language’, and how we relate our thought and thinking with them. Our conceptual thinking takes place in a mental representation of language-like structure; however, this is not direct representation of spoken language. Rather, this ‘language of thought’ is a form of representation where the human mind perceives concepts and symbolizes them rather than using words for those concepts. Hence, the mind uses internal forms of representation, different from traditional linguistic ‘words’, to represent these concepts. Still, these symbols combine to form mental sentences, called the propositional attitudes, which are consistent with the grammatical rules of language. This conceptual thinking has a computational nature that extracts thinking from processing the series of mental symbols according to algorithms. Representational Theory Jerry Fodor specifically implements representational theory in supporting his position (Fodor 1987). According to representational theory, thinking occurs in the form of symbols that are actually the propositional attitudes described earlier (Fodor 1987). While Fodor argues that propositional attitudes should be represented not only as symbols but also as a language, he believes this ‘language of thought’ is different from spoken and written languages like English, French or German (Fodor 1987). Consider the following: ‘I don’t want to eat ladyfingers, so I’d better tell mom to make me French fries.’ According to representational theory, there is a state of a section of this individual’s brain that represents his or her unwillingness to eat ladyfingers (Fodor 1987). There is another section of this individual’s brain that represents his or her way of avoiding ladyfingers; namely, to tell their mom to cook something else (Fodor 1987). Thirdly, there is a small piece of brain circuitry that is linking these two states and instigating an action (Fodor 1987). LOTH asserts that the representation of the decision in this example here has to be structured. That is, it has to be structured just like a sentence articulating that decision. Fodor’s LOTH further states that the structure of symbolic representation in the mind and the linguistic representation of that symbolic representation must be related. That is, the structure of the brain state matching the decision of eating French fries will be equivalent to the structure of the sentence articulating the decision. One considers another example: ‘There is a suspicion crossing my mind that a rat is under my bed’. The ‘suspicion’ is expressed in a sentence: ‘A rat is under my bed.’ According to LOTH, there is a physical state inside the brain that gives rise to this suspicion. This physical state is further divisible into sub-states. Rat is one sub-state, bed is another sub-state, and the other words that are parts of the sentence are other sub-states. The decision of eating French fries and the suspicion of having a rat under the bed is what Fodor calls beliefs. An individual has a belief that the best way to avoid ladyfingers is to ask his or her mom to cook something else. This individual has a belief that a rat is under their bed. According to Fodor, beliefs are structured in our minds in a way sentences are structured. Combinatorial Syntax and Compositional Semantics According to Fodor and Pylyshyn, for the representational handling of symbols to have a linguistic structure, it must utilize both a combinatorial syntax and a compositional semantics (Fodor and Pylyshyn 1988). The representational handling of symbols implements combinatorial syntax if: (1) it utilizes atomic representations and compound representations; and (2) compound representations are composed of either compound or atomic constituents. Moreover, the representational handling of symbols implements compositional semantics if the semantic content of a representation becomes a function of the semantic content of its: (1) syntactic components, (2) the general structure of the representation, and (3) the organization of the components within the structure on the whole. According to Fodor’s LOTH, mental representation of symbols possesses both a combinatorial syntax and a compositional semantics. This means that thoughts that take place in the mind in the form of mental language -- the language of thought. Thoughts are nothing but sentences forming in the mind. There are a number of further considerations that Fodor establishes. Sentences can be expressed through numerous types of expressions or media. For example, they can be written down on paper in the form of formal languages like English and French; they can be encoded and decoded in codes; they can be expressed in computer languages like C, Java, and CSharp; and they can be engraved in stone. Fodor claims these sentences can be interpreted in the brain through encoding at an incredibly elevated level of abstraction. Fodor also suggests that the symbolic representation of symbols in the brain, or the language of thought, is not analogous to the spoken or written languages, like English and German (Fodor 1975). The language of thought is the general linguistic structure of thought. Hence, Fodor suggests that there must be a mental language for the expression of thought in order to learn a spoken language. The language of thought, as asserted by Fodor, is not thoughtfully accessible to the thinker. That is, although the thinker has full control over the constituents of thought -- symbols, images, and other visual patterns -- the linguistic structure is not visible to him or her. Hence, the language of thought can be understood as unconscious processing. Criticisms of Fodor While Fodor’s language of thought has gained great support, a number of prominent criticisms have been levied against his theories. One prominent criticism has targeted computational models. This perspective contends that computational models do not presume a linguistic medium of representation, and instead they are connectionist networks; this is a severe threat to Fodor’s argument for a language of thought. Connectionist networks became popular in the 1980s making some philosophers claim that rational behavior can be explained through these networks, but Fodor and Pylyshyn opposed the idea empirically and argued that the linguistic medium of mental representations was the only plausible theory that could explain rational behavior (Fodor and Pylyshyn 1988). Another prominent criticism of Fodor’s language of thought is that it fails to explain the notion of language learning and, therefore, is subject to an infinite regress. These criticisms consider that since humans are not born with language and must learn it through socialization, so one must also learn the language of thought. Fodor counters this perspective by arguing that the language of thought is unique in that it is innate to human existence. Still, a further criticism is that levied by Daniel Dennett. Dennett argues that Fodor mistakes human instinct for ‘language’. While such a perspective holds weight, one considers that it may be semantics to argue the difference between instinctual understanding and linguistic. Is Fodor right? While there are a number of convincing objections to Fodor’s position, it ultimately appears that Fodor has a strongly establishment argument. Fodor’s argument for a LOTH was an extension of the claim that models of cognition are computational models since they presume a linguistic medium of representation (Fodor 1975). Fodor also states in his 1975 LOTH argument that “remotely plausible theories” are better than having no theories at all to describe mental representations (Fodor 1975). Fodor’s LOTH argument then was the only plausible theory that explained rational behavior or mental representations of symbols by way of a linguistic structure. Productivity Another prominent element supporting Fodor’s perspective is the notion of productivity. Productivity is that characteristic of the system of representations that enables it to fabricate an infinite number of unique representations. Each of the letters in the English alphabet (A, B, C,…) is a unique atomic representation and is capable of producing an infinite number of sentences, making the representational system productive. In contrast, a button on the switchboard has only two atomic representations: on and off. This representational system is non-productive. Hence, we see that productivity occurs when a finite number of atomic representations combine to form an infinite number of compound representations with unlimited length through the implementation of combinatorial syntax and compositional semantics. Fodor and Pylyshyn assert that mental representations of thoughts are productive because the system of representations is expressed by both combinatorial syntax and compositional semantics (Fodor and Pylyshyn 1988). Natural languages tend to be productive because a finite number of alphabets combine to form an infinite number of sentences with no maximum value of length. Hence, competent speakers of a native language are able to construct and understand countless distinct sentences that make the system of representations productive. In the context of LOTH, one is able to construct and comprehend an infinite number of exclusive thoughts. Still, “this unbounded expressive power must presumably be achieved by finite means” (Fodor and Pylyshyn 1988, p.33). This indicates that humans do not hold unique atomic mental representations since humans construct an infinite number of thoughts using a finite number of atomic mental representations. This is why the human mental representational systems possess combinatorial syntax and compositional semantics. Hence, Fodor is right in claiming that the linguistic structure of the system of mental representations must utilize both combinatorial syntax and compositional semantics. Systematicity Systematicity is that property of the system of representations that enables it to express propositions in combination with the expression of certain other propositions. For example, ‘I am young and you are old’ and ‘you are old and I am young’. The sentential logic in these two sentences is systematic according to propositions because it can explain the first part of the sentence if and only if it can explain the latter part of the same sentence. Fodor and Pylyshyn argue that thought is systematic because, again, the mental system of representations possesses both the combinatorial syntax and compositional semantics (Fodor and Pylyshyn 1988). Hence, ones see that what makes the two propositions systematically related to each other is that those propositions should both be expressed as compound representations within the representational system with the same structure and components of structure. The propositions may differ only in the order of components. This is why the propositions ‘I am young and you are old’ and ‘you are old and I am young’ are systematically related to each other. The representation of the first proposition is ‘I.Y’ and that of the second is ‘Y.I’. As such, they contain conjunctions with the same structural components, and they only disagree in the order of components within the structure. Systematicity contends that all such representational systems possess both the combinatorial syntax and compositional semantics. Hence, systematicity refers to the fact that all mental representations have linguistic structure, as Fodor’s LOTH asserts. Ultimately, such a notion presents a near objective articulation of thought as language. Inferential Coherence Another strong element supporting Fodor’s perspective is inferential coherence. Rowlands argues that “Inferential Coherence refers to the capacity of the thinker to draw inferences based on the structure of the thought she entertains” (Rowlands, 1997, p. 22). For example, ‘Bob and Michelle are going to get married’. Two statements can be inferred from this: ‘Bob is going to get married’ and ‘Michelle is going to get married’. Hence, a logical inference is required to make a system inferentially coherent. If Y implies ‘Bob and Michelle are going to get married’, and Z implies ‘Bob is going to get married’, then Y is the logical conjunction and Z is the first conjunct of Y. The syntactic structure of the representations seen in such examples reflects the semantic structure of the propositions being represented. Fodor refers to inferential coherence as a solid defense supporting the hypothesis that mental representations have a linguistic structure. Similar to the previous examples, Fodor’s position regarding inferential coherence is highly logical and consistent. Conclusion Ultimately, Fodor’s LOTH articulates how inner mental representations can be termed as thoughts, and how thoughts are represented in the human brain. The essay evaluated Fodor’s hypothesis regarding why there must be a language of thought, supporting his position. Thoughts were examined as mental sentences formed of components known as mental words that are only visualized in the mind. However, the essay established that the language of thought is different from natural language in that it is a mental representation of symbols and images. The concepts of propositional attitudes, combinatorial syntax, compositional semantics, productivity, systematicity, and inferential coherence further explain and verify the only plausible theory of the language of thought. References Fodor, J.A., (1975). The Language of Thought (Harvard University Press, Cambridge, 1975). Fodor, J.A., (1987). Psychosemantics: The Problem of Meaning in the Philosophy of Mind (Cambridge MS: MIT Press, 1987). Fodor, J.A., (1987). ‘Why there still has to be a language of thought’, in Psychosomatics: the Problem of Meaning in the Philosophy of Mind (MIT Press, Cambridge, 1987), pp. 136-154. Fodor, J.A. & Pylyshyn, Z.W., (1988). ‘Connectionism and cognitive architecture: a critical analysis’, Cognition, vol. 28, issue 1-2, 1988, pp. 3-71. Rowlands, M., (1999). The Body in Mind: Understanding Cognitive Processes (Cambridge University Press, USA, 1999), p. 27. Read More