In this essay I will be outlining arguments both for and against the existence of propositional attitudes, that is, the existence of mental states that people possess that have a relation to some proposition. Certain mental states such as believing x and desiring y all follow the structure that one would see from an outlined proposition, and there is much room for discussion on if these exist as cognitive states. Paul Churchland (1981) is a key figure in this topic, arguing that propositional attitudes do not exist and are instead part of a common-sense “Folk Psychology” that will inevitably be replaced by a more convincing theory – claimed by Churchland to be Neuroscience. Section 1 is dedicated to his argument in full. Section 2 is dedicated to responses made in defence of propositional attitudes which target the consistency or accuracy of Churchland’s “Eliminative Materialism”. I go on to argue that propositional attitudes do exist by virtue of them being a part of an innate linguistic framework, making reference to Jerry Fodor’s “Representational Theory of Mind”.
Eliminative Materialism is the view that our intuitive, common-sense conception of the mind and mental states is so fundamentally mistaken that it will ultimately not be refined into something better but done away with entirely and replaced with a superior model, namely a completed form of neuroscience (Churchland, 1981, p.67). This is argued by way of claiming that this common sense view of mental states, henceforth known as folk psychology (or FP), is a faulty theory much like alchemy or phlogiston that ought to be rejected, and because of this the theory’s notion of propositional attitudes such as beliefs and desires do not really exist.
First and foremost, how should FP be considered a theory? There is a definitive structure to how theories are arranged; they consist of basic laws that are assumed by adherents of the theory, and apply to entities that are assumed to exist just as much as the theory’s laws are. Those laws and entities are then appealed to in order to explain an event or to predict future events (ibid, p.69). Churchland argues that FP is no exception – we all tacitly assume that there are relations between mental states and the external environment that come to form our description of the behaviour of people external to us (ibid.). This “body of lore” that we refer to is identical to theories in the past by virtue of it making an appeal to semantic terms, the meanings of such terms appealing to a set of assumed laws that may or may not be true. If it can be proven that these laws are not entirely true, the theory in which we are using them must be revised or even rejected, which is what Churchland holds for FP. He does concede that FP solves a lot of problems, namely that of the problem of other minds. FP’s assumption that external minds exist means that this problem is not something that needs to be considered for the theory itself to work – it already assumes that this is the case (ibid.). That said, this is more an appeal of the theory and not an outright solving of the issue. An assumption of this sort brushes the problem of other minds aside to avoid the trouble of accounting for it, which is what subtly leads to Churchland’s belief that the theory of FP is not accepted for its standing as a good theory.
With this in mind, what faults does FP have that make it a bad theory? A key element of FP’s failures is its inability to explain various psychological phenomena. Churchland points to examples like the complexity of mental illness, how the imagination and human creativity works, and the dynamics of sleep (ibid, p.73). All of these things are not inherently reflective of propositional attitudes, and it seems like FP is not able to explain their interaction with the rest of the mind – it can only go so far. Furthermore, given its status as a common-sense theory, despite the fact that FP has existed at least as an intuition for hundreds of years, it has not seemed to make the same kind of progress that successful theories make over time. The theory of gravity, for example, tends to be refined with broader physical paradigms in its lifespan – Newtonian, Einsteinian and presumably soon Quantum physics - and as a result it has persisted because of its status as a successful theory. The same cannot be said for FP; Churchland claims that its life has been “one of retreat, infertility, and decadence.” (ibid, p.74).
On top of these problems is the problem of FP’s lack of integration with other theories. This was touched upon somewhat in the previous point – conceptions of gravity end up integrating with a more fundamental theory over time, and as such its effectiveness as a theory is in its successful integration with scientific advancements. FP does not share this trait, its laws and ontology have no bearing or relation to other fundamental disciplines, it is a theory that, according to Churchland, is impossible to unify with others (ibid, p. 75).
After outlining the faults with FP, Churchland can make the case that its ineffectiveness as a theory is what warrants our replacing it. It is too far gone given the unification of other theories and their perceived effectiveness to try and salvage in any way, and should only exist in the history books in the same way theories of alchemy were before atomic theory was more fully developed thanks to advances in physics. FP, to Churchland, will inevitably be scrubbed away in place of the successes of neuroscience.
Now that eliminative materialism’s cards are on the table, what hope is there to defend it? The structure of Churchland’s argument is fairly evident: FP is a theory, and a bad one at that, and because of this it must be replaced by one that is better, such as a completed neuroscience. In this way, we can challenge this line of reasoning by attacking its individual pieces. However, I believe that Churchland is right in saying that FP is a theory. People that object to this idea include Barbara Hannan (1993), who argues Churchland is mistaken in assuming FP serves the same purpose as any empirically-minded scientific theory, as it means to reflect an individual’s capacity for inductive reasoning, and therefore does not require “scientific vindication” (Hannan, 1993, p.169). In other words, FP does not have to improve scientific progress with regards to the mind as it is not part of that union, and this is an idea that will be returned to later.
More to the point, I believe the real issue in Churchland’s reasoning is his attack on functionalism. Frank Jackson and Philip Pettit (1990) make the case that his conception of FP aligns with a “commonsense version of functionalism as applied most particularly to beliefs and desires” (Jackson and Pettit, 1990, p. 33). By this, they mean that Churchland’s view of FP is identical to the relations of external stimuli to internal states and therefore external behaviour. Churchland acknowledges this by his choice of language when describing FP as a theory: “the semantics of the terms in our familiar mentalistic vocabulary is to be understood in the same manner as theoretical terms generally” (Churchland, 1981, p. 68).
Jerry Fodor’s physicalist notions are something that ought to appeal to Churchland’s attempt at a unification of the sciences sans folk psychological concepts, but Fodor would argue that the predicates neuroscience requires by virtue of being a “special science” do not square with the “generality of physics” (Fodor, 1974, p. 97). Every event attributed to a special science predicate - such as belief states being attributed to the firing of particular neurons - also falls under events attributed to physical science predicates, but not vice versa.
Eliminative materialism is an inherently reductionist theory, and Fodor reasons that reductionists require “bridge laws” that use identity statements to link special sciences - such as neuroscience - with fundamental physics. Such a move is disjunctive as physics is too generalised to require bridge laws, and therefore the special sciences are inherently unable to unify with it (ibid., pp. 103-104). In this sense, Churchland’s putting-down of FP to promote the unification of neuroscience is based on the false assumption that neuroscience is part of the integration he seeks out of an explanation for mental states. Because of this, the multiple realizability offered by functionalism is especially appealing when talking about mental states due to the fact that these states can be realised in various kinds of physical systems, and so avoids the drawbacks involved in using a special science to do so. Fodor’s claims defend FP, as he believes them to share features with a matured form of psychology, those features being intentionality and causality (Lloyd, 1991, p. 289). Contrary to Churchland’s assumptions, Fodor merely believes that the folk-psychological conditions for propositional attitudes such as belief are not fit for an empirical type of science, but exist nonetheless through the so-called “law-like relations” of these systems (ibid.).
One issue remains: if FP can be protected from eliminative materialism by virtue of its functional relations between causal stimuli, internal states and external outputs, how can we talk of the intentional and causal states of the mind in physical terms? Therein lies Fodor’s Representational Theory of Mind. Simply put, our mental processes are inherently computational, and states of intentionality (such as propositional attitudes) are relations to representations that all mental states take place within. These representations, the meanings of which are derived from our internal states, are semantic in nature, and are physical in nature by virtue of the proven existence of mechanical devices capable of making inferences from the meanings associated with various states. Fodor is concerned here with the preservation of truth in detached internal states, and he argues that the logical relations between these states is based on the syntax of this mental language we abide by, the “body of lore” that allows our minds to acknowledge our own states and communicate the states of others to ourselves. Thinking is both rational and mechanical, and in this way propositional attitudes have actual truth value. To have a belief that something is the case is to have a logic-abiding, computational relation to a state that has the meaning of that belief (Lloyd, 1991, p. 290).
Throughout this essay I have outlined key arguments against the existence of propositional attitudes, and fought against them using arguments from more convincing theories. I have then asserted the existence of propositional attitudes by way of positing their necessity within an internal language of thought, and believe this to hold much more water than the perceived need to do away with propositional attitudes entirely.
Hannan, B. (1993). “Don’t Stop Believing: The Case Against Eliminative Materialism.” Mind & Language 8(2), pp. 165-178.
Lloyd, D. (1991). [Review of Psychosemantics: The Problem of Meaning in the Philosophy of Mind, by J. A. Fodor]. The Philosophical Review, 100(2), pp. 289–293.
Jackson, F. and Pettit, P. (1990). "In Defence of Folk Psychology." Philosophical Studies, 59(1), pp. 31-54.
Fodor, J. A. (1974). Special Sciences (Or: The Disunity of Science as a Working Hypothesis). Synthese, 28(2), pp. 97–115.
Churchland, P. (1981). “Eliminative Materialism and the Propositional Attitudes.” The Journal of Philosophy 48(2), pp. 67-90.