Causal Theories of Mental Content

by Dr. Charles Wallis
Last Revision 4/21/2012

Theories

Nomic Covariation

Advantages of Covariation Theories

Covariation Theories and the Disjunction Problem

Functional Role Semantics

Advantages of Functional Role Semantics

Problems for Functional Role Semantics

Causal Theories and Cognitive Science

Bibliography

Other Internet Resources

Related Articles

Introduction 

It is a platitude, denied by very few, that minds causally interact with the world.  Causal theories of mental content suppose that such causal connections not only facilitate the mind's interactions with the world, they provide the basis of mentality.  A large part of understanding the nature of the mind is understanding how it comes to be aware of the world.  Causal theories develop in the context of some or other representational theory of mind (RTM).  A representational theory of mind holds that mental states have intentionality (are about the world) in virtue of a representational relationship holding between the mental state and the object.  Causal theories of mental content hold that mental states represent the world in virtue of the sorts of causal relationships those states have--causal relationships with other states within the mind and/or causal relationships between the mental state and the world.  Historically, causal theories have been attractive to physicalists; philosophers interested in understanding all mental properties, states, and/or events (and all of existence) in terms of physical properties, objects, relations, states, and/or events.  Such naturalistic philosophers seek to understand the mind as purely physical in origin and nature.  The main body of this article discusses the historical precedents of contemporary causal theories of mental content, the two major types of causal theories of mental content current in the philosophical literature.  In each case the article outlines the theory, discusses its advantages and strengths, raises commonly perceived problems for the theory, and presents responses by proponents of the theory.  

Historical Precedents

Though many theories of mental content are causal theories, not all theories of mental representation are causal in nature.  For example, Locke supposes that ideas of primary qualities represented qualities in an object, not through reliable causation, but because they were similar (share the same properties). Contemporary causal theories of mental content have predecessors dating back to Book III of John Locke’s Essay Concerning Human Understanding, possibly Aristotle’s De Anima, or even to Plato’s Theatetus.  Locke’s notion of secondary qualities (a quality or power of the object to cause particular ideas in us which bear no similarity to the object, e.x., color) looks very much like a contemporary causal theory based upon reliable causally mediated covariation  (Cummins 1989).  Aristotle’s discussion in Book II of De Anima suggests that, for example, color in the object is different from our sensations of color, which are nevertheless reliably caused by light hitting the object.  Similarly, Plato’s analogy of perception as the matching of sensations caused by the world to impressions (knowledge) upon a ball of wax also suggests that Plato entertained the notion that causal connections allowed our minds to represent.   Contemporary causal theories of  mental representation have developed in the theoretical context of explanation in cognitive science.  It can be helpful to view the intended role of such causal theories in computation explanations in cognitive science.  However, contemporary causal theories can be understood independently of their role in cognitive science.  Sidebar on Explanations in Cognitive Science

Theories
There are two general causal theoretical approaches to a theory of representation in the contemporary literature.  Neither theoretical approach regarding the representation relation has gained general acceptance.   However, there are several agreed-upon constraints for a theory of representation.  First, the representation relation must be consistent with the physicalistic nature of cognitive science and science generally.  Second, the relation must be present and explanatory in accepted explanations within cognitive science (an appropriate characterization of "accepted" explanations is somewhat controversial).  Philosophers particularly would also like to see the accepted theory of representation explain many of the widely accepted properties of belief.  Block (1986) and Fodor (1988) provide excellent discussions of the potential problems, the solutions to which many philosophers hope to find in a theory of mental representation. 

One theoretic approach to articulating the representation relation, "Nomic (law-like) covariation" or "covariation," postulates a simple causal relationship between the object, property, event, or relation and the state that represents the object, property, event, or relation.  The other approach, "functional role semantics," hypothesizes that a state has content in virtue of the state’s occupying a particular position in a complex web of causal relationships characterizing the cognizer’s functioning.  Candidate content-fixing causal relationships include causal relationships within the cognizer (i.e., relations between brain states) and/or outside the cognizer (i.e., relations to the distal environment).  Both functional role semantics and covariation theories satisfy the (above) first, physicalistic constraint by hypothesizing that a cognizer’s states represent the distal environment solely as a result of the specific sorts of causal connections had by those states.  The theories diverge in terms of the specific causal relationships each emphasizes.

Covariation
Covariation theories hypothesize a state represents an object or property in virtue of a causal connection between the object, property, event, or relation in the world and the state that represents that object, property, event, or relation within the cognizer.   Information theory as well as work within a psychophysics inspire contemporary covariance theorists.  Fodor, for instance, notes that 

Covariation theories, as a result, seem to satisfy the second constraint upon theories of representation, i.e., that the relation must be present and explanatory in accepted explanations within cognitive science.  For example, Hubel and Wiesel (1977) investigate the representational content of cells in the striate (visual) cortex by monitoring the activity of those cells looking for preferential relationships between the activity of these cells and the presence of properties in the visual field.   A system’s states represent those objects, properties, events, and/or relations of the distal world with which they covary according to covariation theories.  Specifically, covariation theories assign contents to states via some version of the following definitions:    

    Illustrative animation

Advantages of Covariation Theories
Intuitive Appeal
The intuitive appeal of covariation theories stems from the idea that a cognizer represents an object, property, event, or relation by being causally "in tune" with that object, property, event, or relation.  The causal connection provides a concrete connection between the object, property, event, or relation and the state which represents it.  It also insures that the state represents elements of the cognizer's distal environment.  

Punctate Content

Further, covariationists such as Jerry Fodor advocate the approach because covariation assigns content to individual states independent of how the cognizer operates upon those states (inferences it can make) and/or the content of other states with which the state might interact.  Fodor refers to this assignment of content as "Punctate Content," or "atomic content."  Among the advantages claimed by Fodor for punctate content is that it allows for content identity across individuals who have quite different theories regarding some object, property, event, or relation.  Similarly, punctate content provides the only theory whereby people can refer to real objects, properties, events, or relations even when they have great numbers of false beliefs about them.  For instance, Fodor claims (1992) that Aristotle thought about and talked about the same things we refer to by "stars" even though he falsely believed stars rotate relatively closely around the Earth on glassy spheres.

Systematicity and Compositionality

The punctate content argument for covariation theories supposes that it is an advantage that the content of representational states is fixed in isolation from inferences in which the state might participate.  In contrast, the systematicity and compositionality argument suppose that covariation theories better explain certain perceived truths about the combinatorial properties of language an of concepts.  Compositionality asserts that the meaning of a complex expression in a language results from the meanings of its constitutive elements.   Compositionality plays an central role in many linguistic theories since assuming compositionality, for both language and thought, provides one with a fairly straightforward explanation of the human ability to grasp an enormous number of different thoughts of varying complexity and their corresponding linguistic expressions.  For instance, because we understand the individual elements like "cup" and "coffee", we understand the complex expressions "cup of coffee," "hot coffee," "coffee gives me the shakes," etc..  One explains our understanding in each case by noting that the meaning of these complex sentences results from the meaning of their constitutive elements.  Furthermore, languages and thoughts seem to have a systematic structure to their compositionality.  For instance, one seemingly cannot have the thought that "coffee has more kick that tea" without also being capable of thinking that "tea has more kick than coffee."   Proponents of covariation theories (Fodor and Pylyshyn 1988, Fodor and McLaughlin 1991) argue that such seeming facts about thought and language are naturally explained by covariationist semantics.

Epistemology and Representation

A final perceived advantage for covariation approaches lies in that they provide a clear-cut mechanism through which a cognizer can come to know about an object, property, event, or relation in the world in virtue of their representational capacities.  A cognizer that represents an object, property, event, or relation in the environment does so insofar as the cognizer can reliably detect the presence of the object, property, event, or relation.  (Dretske, 1989)
 

Covariation Theories and the Disjunction Problem
Theorists call the most widely discussed problem for covariation theories the "Disjunction Problem".  According to covariationists, a state represents an object, property, event, or relation if the system tokens (enters into) that state when confronted with the object, property, event, or relation and only when confronted with that object, property, event, or relation. However, suppose that cognizer tokens a state, S_c, in all of its cat interactions.  According to the when clause of the covariationist definition, the cognizer’s state appears to represent cathood. Unfortunately, one night the cognizer tokens an S_c when seeing a skunk in the darkness. The only when clause of the definition prima facie dictates that S_c never really represented cathood simpliciter.  Rather, the cognizer has always represented a property that we can only describe using the disjunction, "cat or skunk".

Illustrative Animation

The disjunction problem poses difficulties for covariation theories in two ways: First, the theory seems to dictate counterintuitive representational contents for states.  If one has beliefs about cats, (e.x. cats are domesticated felines that have a number of distinct breeds), then covariation seems to dictate that those beliefs have always been about things having the property a cat or a skunk. Second, the disjunction problem seems to demonstrate that covariation cannot account for misrepresentation, since any seeming case of misrepresentation by a cognizer becomes a correct representation of a disjunctive property under covariation.  In other words, one never mistakenly believes that one sees one's mother at the corner, one always correctly believes that "mother/other" is at the corner.

Theorists have explored a number of solutions to the disjunction problem.  All of these solutions rely upon some form of the idealization strategy: A given solution will separate cases of the tokening (occurrence) of a state into two groups, one in which content is already fixed and representational error can occur (normal conditions), and the second class (ideal circumstances) in which perfect covariation fixes content. Several important variations of this solution have evolved in the literature.

The Two-Cases Solutions to the Disjunction Problem
To solve the disjunction problem, one must modify basic covariation theory so that the cognizer represents "cat" and not "cat/skunk" despite occasionally mistaking skunks for cats.  Theorists usually adopt the same general solution strategy; I call this strategy the two-cases solution strategy.  According to the two-cases solution strategy tokenings of S_c fall into two classes--content-imbuing and content-utilizing classes.  In the content-imbuing classes strict nomic covariation holds and the cognizer never falsely tokens an S_c.  In the content-utilizing cases, the cognizer can falsely token S_cs.  The false tokenings of S_c in the content-utilizing cases misrepresent because content remains fixed from the content-imbuing cases.  Thus, the two-cases solution only works insofar as its advocate can plausibly differentiate content-imbuing and content-utilizing cases and in such as way that content-imbuing cases plausibly fix representational content. Theorist offer several versions of the two-case solution in the mental representation literature.  The various versions of the two-case solution strategy differ with regards to the account their advocates offer for distinguishing content-imbuing from content-utilizing cases.        

Ideal Conditions
Advocates of the ideal conditions solution to the disjunction problem hold that under ideal perceptual conditions (i.e., in good light, at close distance, etc.) one can distinguish cats from skunks. Thus, ideal condition theorists suggest, cognizers represent cats as cats because of their discriminatory capacities under ideal circumstances.  Idealization in mainstream science both inspires the covariationists' idealization strategy and provide the source of the theory's tacit plausibility.  In normal cases of scientific idealization theorists ignore certain negligible parameters in real systems to formulate a law. The resulting law, though not strictly true of any actual system, proves predicatively adequate, and quantifies a real relation in actual physical systems.  For example, real gases are not composed of Newtonian molecules.  Real gas molecules are not point masses, nor are their collisions perfectly elastic.  Likewise, heat is not the only energy source present in gases.  In most cases, however, other parameters (like electromagnetic forces) prove negligible.  One idealizes away from such negligible parameters to quantify the inverse relationship between, on the one side, the pressure and volume of a gas, and on the other side, the temperature and number of moles.  Idealizing away from other parameters results in the ideal gas law: PV = nRT.

The move toward idealization in response to the disjunction problem attempts to emulate noncontroversial cases of scientific idealization like the above-described ideal gas law. The basic idealization move defines representation as follows:

On the idealization line, S_c represents cats as cats since, under ideal circumstances, the cognizer tokens S_cs when, only when, and because of the presence of instances of cathood.  State S_c does not represent "cat or skunk" because under ideal circumstances skunks do not cause the cognizer to token S_c.  Such an idealization from error, admits the covariationist, breaks down in abnormal circumstances.  Nevertheless, the idealization allows psychological laws to capture a real relation (the representation relation) in actual systems.  Moreover, since most cases are close to ideal circumstances, psychological laws utilizing idealized covariation prove predicatively adequate.

The covariationist’s ideal conditions solution divides tokenings of a state by a system into content-imbuing (ideal) and content-utilizing (malfunctioning/atypical) classes.  Covariationists assert that one legitimately idealizes away from error because errors prove coextensive with the cases of malfunction and/or atypical situations (from which one legitimately idealizes).  For example, if one pulls a weighted spring hard enough to bend or break the spring, it will no longer obey the standard harmonic oscillation equation.  Such spring systems do not undermine the the basic physics of dampened harmonic oscillatory systems because the circumstances prevent the spring system from operating qua spring system.  The same holds for cognitive malfunctions: One’s tokening of a S_c as a result of a amphetamine induced psychosis does not introduce disjunctive content because amphetamine induced psychoses count as a case of mental malfunction.

Unfortunately, all skunk-caused instances of S_c need not trace their origins to malfunctions.  One could token a S_c as a result of the same cat-like features prompting tokenings of S_c for cats.  Dismissing such look-alike cases as malfunctioning requires one also dismiss cat cases as malfunctions.  All cognition becomes malfunction, if one focuses exclusively upon malfunction in specifying ideal conditions.  Consider another example.
 

Illustrative Animation

Subjects normally judge the right side of the above figure to have a much darker shade of purple than the left side.  In fact, the surface reflectance of the purple areas of the figure remains uniform (it is all the same color of purple).  Psychologists refer to the mistaken judgment as "the assimilation effect."  Psychologists explain the assimilation effect by attributing the mistake to the pooling together of signals from several retinal cells.  In most circumstances, pooling reduces equivocation (error or noise) due to false signals from individual retinal cells. However, when one closely intersperses the dark and light patterns throughout a visual field, pooling serves to inhibit or to excite cell firing, resulting in a misrepresentation of relative shading.  So, a case of normal functioning--in fact, functioning that acts to reduce error--results in misrepresentation.

Covariance theorists respond to such normal functioning cases by appealing to atypical conditions.  For instance, when the pressure upon a gas exceeds atmospheric level (1.013 x 105 N/m2), or when its temperature becomes too great, the negligible parameters ignored by the idealized gas law become non-negligible.  The predictive accuracy of the idealized gas law plummets, though the gas does not malfunction.  Nevertheless, these cases do not count against the idealized gas law because the circumstances are atypical.  [These cases are actually quite common in the universe at large, but these cases count as atypical since they violate the ideal conditions] The covariationist claims a similar atypicality of perceptual circumstances in assimilation cases.  In short, skunk-caused tokenings of S_c--whether malfunctions or atypical cases--count as less than ideal.  As such, these cases have no effect upon the content of S_c in cases when the system tokens a S_c in response to a cat.

The covariationists’ move looks suspiciously circular unless they can specify an independent means of ruling-out cases of similar features as atypical.  In other words, one must define ideal conditions so that malfunction and atypical circumstances prove coextensive with--capture all cases of--error.  But one cannot avail oneself of the notion of error, nor of other intentional or semantic notions in formulating and motivating the definition.  One must have a reason for labeling the skunk to S_c cases as less than ideal, and the reason cannot be that skunks cause cat tokens (S_c) in feature similarity cases.  For example, in the movie "The Crying Game" an Irishman trying to escape from the IRA becomes romantically involved with a woman.  However, much to the surprise of the Irish man (and the audience) this woman is actually a man.  The reaction of the  Irishman clearly shows he misrepresented the gender of his romantic partner.  Yet, the Irishman did see his partner in good light, at close distance, etc..  The natural solution to this difficulty attempts to take advantage of the idea that there are conditions under which such a sex difference would not escape notice.  However, in specifying such conditions as ideal conditions for this case, one’s knowledge of the property that the state actually represents seemingly guides one's specification of ideal conditions--rendering the specification circular.  That is, the ideal conditions theorist's conditions seem to capture rather than explain content since the ideal conditions theorist must know the representational content in order to correctly and completely specify the ideal conditions.

Additionally, critics (Wallis 1994 and 1994a) argue that appealing to idealization to defeat the disjunction problem proves strongly and negatively disanalogous with successful uses of idealization in science.  For example, in order to explain misrepresentation in cases of malfunction or atypical circumstances, the covariationist must rely upon the content dictated by idealized nomic covariation.  When a skunk causes a S_c because of bad lighting, the covariationist must suppose that S_c represents cats in order to explain why the cognizer misrepresented the skunk as a cat.  Thus, for the covariationist, the results of the idealization apply in conditions which violate the presuppositions of the idealization.  In contrast, when the pressure upon a gas exceeds atmospheric level (1.013 x 105 N/m2), or when its temperature becomes too great, the negligible parameters ignored by the idealized gas law become non-negligible.  In these atypical circumstances physicists will agree that the ideal gas law does not apply since the presuppositions of the idealization do not hold.  Thus, while the mainstream use of idealization limits the use of idealized laws to ideal or close to ideal circumstances, the covariationist cannot accept such limitations.  Indeed, their strategy must deny those limitations.

Learning Periods
Fred Dretske (1981, 1989) suggests an alternative account of content-imbuing and content-utilizing classes.  Dretske suggests that content gets fixed for a state during a learning period.  In this period, cognizer develops a perfect causal connection between the state and the object or property with the help of an instructor who provides examples and corrections.  Once the learning period ends, the cognizer’s state has a fixed content and tokenings of the state in cases in the absence of the  object, property, event, or relation count as misrepresentations--content-utilization cases.

Or, to adopt probabilistic formulation more consistent with Dretske's work:

Dretske motivates his learning-period-based content-imbuing versus content-utilizing distinction by appealing to the role of learning in cognition.  The role, on this account of learning, of learning consists in recruiting states for specific representational purposes.  The learning period recruits and fine-tunes the covariation between a state and the object, property, event, or relation.  Once learning fixes the covariation, then the system utilizes the recruited state to represent that object, property, event, or relation. 

Fodor (1988) and others criticize the learning period approach on two grounds.  First, there seems to be no principled distinction between learning and non-learning periods.  Hence, theorists lack adequate grounds for calling some tokenings of the state content-imbuing and others utilizing (i.e., representing or misrepresenting).  Second, even if one could specify a learning period, learning theorists seem to lack a principled distinction between the univocal (single) and disjunctive causal connections.  Critics ask why, if one tokens S_c for both cat and skunk in the post-learning period, shouldn’t one suppose that the causal connection created in the learning period is between cat and S_c and not between cat/skunk and S_c?

Philosophers also offer the following standard criticisms of the learning periods approach.  First, a cognizer can prima facie only represent those objects/properties for which it has had a learning situation.  Thus, the learning period approach cannot accommodate innate knowledge.  Second, cognizers do not seem to learn to identify an object, property, event, or relation with perfect reliability as required by Dretske.  Finally, since learning is a historical fact about a cognizer and a duplicate does not share one's learning history,  the states of one's molecule for molecule duplicate would seem to lack the content had by one's own states.  (Such examples duplicate examples appear in Burge, 1979, Cummins 1989, Davidson 1987, and Putnam 1975).

Teleological Accounts
Advocates of a teleological solution to the disjunction problem, such as  Dretske (1988), Millikan (1983 and 1986),  Neander (1995 and 1996), Papineau (1984), Shapiro (1992 and 1996), and Sterelny (1990) suggest that evolutionary history (evolution or learning for Dretske) determines content.  For example, a frog captures and eats any ambient moving dot in its visual field.  As a result, hungry frogs will eat ball-bearings rolled in front of them in a lab.  Advocates of the teleological account claim that frogs misrepresent ball-bearings as flies because the uses of the frog’s visual cells responsible for a frog’s ability to propagate its genome, i.e., uses involving fly capturing, determine the function of those cells.  Thus, cases operant in frog evolution (i.e., cases operant in the evolutionary selection of frogs) constitute content-imbuing cases.  All other cases constitute the content-utilization class.  Such an account defines representation as follows:

There is one caveat to the above definition in that notion of representation for teleosemanticists might be more like representing the presence of C at a time.  Since the teleological account depends upon the selectional history of the organism and that history changes over time, the content of the state depends upon the current selectional history.

Much like the learning period theorists, the teleosemanticist distinguishes between content-imbuing and content-utilizing cases based upon the role of evolutionary selection in establishing the nomic covariation between the state and the object, property, event, or relation.  Cases of selection imbue content, whereas all other cases utilize that content.

Two major distinctions appear among teleological theorists: (1) whether function of belief-generating mechanisms generically or of individual beliefs specifically determine the content of mental states, and (2) whether the producers of states or to the consumers of states determine the function (and hence, the content) of those states.  For example, Millikan holds that adaptational history determines the function (and hence, the content) of specific beliefs.  The content of the frog's visual state, fly, has an indicative function, in virtue of the state's past co-occurrence with flies.  Others (ex. Papineau) hold that selection determines the function of belief-generating mechanisms.  Thus, the frog's visual system has the general selective function of indicating the presence of objects in the distal environment.  The particular state means fly because evolution selected a mechanism (the frog's visual system), in part, for its fly detection ability and that mechanism generated that state.  Second, some teleologists  (ex. Millikan) argue that the function of the state for the consumers of the the state determines the content of the state for the producers of that state.  Others suppose that the function of the state form the producers of the state determines the content of that state.  Thus, for Papineau it is the function of the visual system (the producer of the state) that determines its content.  In contrast, the fact that S_c triggers the frog's capturing mechanism (the consumer) that gives S_c its indicative function on Millikan's account.     

Fodor (1988) criticizes teleological accounts on the ground that evolution does not select with sufficient precision to account for typical univocal content claims.  If a frog represents ambient dots as "fly or ball-bearing" in a fly-rich, ball-bearing-poor environment, then that disjunctive representational content would account for the frog’s ability to propagate its genome.  Hence, claims Fodor, evolutionary history will not favor "fly" over "fly-or-ball-bearing" as the content of frog visual cells.  Any contents that would facilitate evolutionary selection with equal efficacy, Fodor claims, have prima facie equal claim as the state's content.

Two other objections to teleological accounts appear in the literature.  First, most artifacts (e.x., compact discs, SUVs, etc.) did not exist during most of the 4.4 million years of hominid evolution.  As a result, advocates of teleological solutions to the disjunction problem must explain why/how a huge percentage of the representational capacities of humans, including the representations of many ordinary objects like chairs or beers, have come into existence after evolutionary selection.  Generally, teleosemanticists suppose these states gain their representational content through those states with primitive, selectional content.  Second, because teleological accounts appeal to the evolutionary history of a cognizer to explain representational abilities, physically and functionally identical states lacking such a history would not have representational content on the teleological account.  For example, if a molecule for molecule duplicate of Millikan where to spontaneously appear, it would seem to have all of her cognitive/representational abilities.  Yet, on Millikan’s account, her exact double would not have any states with representational content.  (Cummins, 1989, Davidson, 1987).

Asymmetric Dependence
Fodor’s (1990) asymmetric dependence theory suggests another twist on the general two-case strategy.  In addition to a nomic connection between a state, S_c, and an object, property, event, or relation like cathood, asymmetric dependence theories define representation in terms of two counterfactual scenarios: 

Illustrative Animation

As in the above versions of the two-case strategy, asymmetric dependence theory assigns potentially troublesome cases to the content-utilization class.  The rational for the division between content-imbuing and content-utilizing classes as well as the assignment of cases to each class lies in the asymmetry between veridical and non-veridical tokenings inherent in the theory.  On the asymmetric dependence account a counterfactual set of conditions exists in which subjects can distinguish, for instance, cats from skunks.  In other words, the theory supposes that the laws of operation themselves provide the basis for the two classes, albeit counterfactual in nature.  The state represents, say, cats and not skunks because cattiness and not skunkiness underlies the operation of the laws.  Critics (Cummins, 1989 and Wallis 1995) argue that asymmetric dependence theories fare no better than other versions of nomic covariation.  One difficulty facing the asymmetric dependence theorist stems from the fact that the brain recognizes higher-level concepts through the detection of lower-level features.  As a consequence, no asymmetric dependence between cases of veridical and non-veridical representation emerge from the laws dictating tokenings, or worse, the asymmetry goes the wrong way--from fake to representation.   For instance, if one finds one's car by looking for features x,y,z, then Fodor can break the car to S_car connection by altering the appearance of one's car.  But that will not break the look-alike-car to S_car connection (violating 2).  If, on the other hand, Fodor breaks the look-alike-car to S_car connection (say by altering one's beliefs about one's car’s appearance), then it seems Fodor breaks the car to S_car connection (violating 1).  Similar stories can be told in terms of the normal, albeit somewhat noisy, functioning of cells in the visual system. (Wallis 1995).

Other philosophers have forwarded other objections.  Adams and Aizawa (1997) argue that Fodor's conditions commit him to the retinal projections of objects, properties, events, and relations instead of the objects, properties, events, and relations themselves.   Antony and Levine (1991) as well Wallis (1995) argue that Fodor's theory runs counter to other aspects of his own account of mentality.  
 

Other Problems for Nomic Covariation
Semantic Reduction and Verificationism
In addition to the disjunction problem, another objection challenges the ability of covariation theories to explain the representational capacities of states with regard to any object, property, event, or relation beyond low-level perceptual properties (e.x. color).  In other words, the objection suggests that one can make one's best case for law-like covariation determining the content of states only for states representing low-level properties like red or zero-crossings.  Even if, one could extend the notion to ordinary, medium sized objects/properties like table or cow, at some point, probably early on, one must suppose that states represent higher-level properties and objects , at least one's like electron, in virtue of their being defined (perhaps even implicitly) in terms of the representational properties of lower level states.

Robert Cummins (1989) objects to this strategy, claiming that the covariation of mental states with objects, properties, events, or relations requires that the cognizer represent heuristic information either implicitly or explicitly.  In other words, nomic covariation with higher-level properties requires a system to operate in accordance with or instantiate a theory.  Such theories require simplifying assumptions to remain tractable.  Simplifying heuristic assumptions, by definition, fail under conditions that violate them.  Thus, these theories never achieve perfect nomic covariation.  As a result, nomic covariation explains, at best, the representational properties of only the most primitive representational states.  All higher-level concepts from dog to democracy must get meaning through definition.  At first glance, it seems plausible to suppose that states that represent high-level theoretical properties like electron get their meaning through definition in terms of lower level properties. After all, such properties and objects often have theoretical definitions. However, the failure of such semantic reductions has shaped the history of western philosophy and mathematics (Quine 1951). The logical positivists, for instance, held that the meaning of terms was just the conditions for their verification (the verificationalist theory of meaning).  However, no generally accepted means of spelling out such conditions emerges in the positivist/empiricist literature.  Covariation’s critics ask, "if the state one uses to represent crow gets its meaning from being defined in terms of lower level properties, then why does one have difficulty in articulating that definition in any but the most superficial sense?"  Psychological research also runs contrary to semantic reductionism in that it indicates that perceptual features, usually accidental and contingent to the target, strongly influence object recognition and categorization.  Likewise, the exact make-up of such features proves somewhat variable between individuals and fluid. (See for example, Anquetil and Jeannerod 2007, Barsalou 1987, Barsalou and Medin 1986, Biederman 1987, Maddox et al. 2010, Malt and  Smith 1984, Medin and  Shoben 1988, Rosch 1975a and 1975b, and Rosch and Mervis 1975)

According to Cummins, mental states achieve covariation with cats because one already has explicitly represented or implicit knowledge about cats (ex. they are domesticated felines). Further, if covariation is not the simple, unmediated causal relationship between the object, property, event, or relation and the state, but rather the result of often complex causal interactions within the system, then this suggests that covariation is really just another instance of its competitor, functional role semantics.  Moreover, it suggests that the tokening of higher-level representations of objects, properties, events, and relations constitutes cognition--not the theoretic primitive from which one can explain cognition.  Perhaps David Marr (1980) expresses Cummins' point the most forcefully when Marr says of vision that,

Failure of Univocal Feature Detection for Low-Level Properties     

Finally, some theorists like Kathleen Akins (1996) (also Churchland and Sejnowski 1992 and Wallis 1995) argue that the various cells performing sensory transduction are not properly characterized as the sorts of feature detectors required by nomic covariation.  Such sensory cells do not respond exclusively to the presence of some particular object or property, but can and do respond to other properties.  These cells, likewise, do not function so as to give an all or nothing response to a given stimulus.  Rather, sensory transduction cells show "selective sensitivity" to properties in that they respond to a wider range of stimulus, but respond more strongly to certain stimuli.  For example, theorists sometimes portray rods and cones in the retina as representing a certain wavelength of light.  The response curves of these cells in fact overlap dramatically.

Thus, not all that dissimilar intensity levels of blue or green light can illicit the same response from a red cone as a particular level of red light intensity.  Additionally, perception is riddled by often useful confabulations by the sensory systems.  For example, the visual system almost immediately pools and averages rod and cone activation.  This pooling and averaging, as noted above, has a good effect in that it eliminates noise from sources like body heat.  However, pooling can also lead to the assimilation effect noted earlier.  Similarly, the visual system compensates for the blind spots in each retina were the cells leave the eye by "filling in" that area.

blind spot

The Problem of Depth

Covariationist intend their definition to capture a representation relation between states of the cognizer and objects and properties in the immediate distal environment.  However, causal covariation occurs at a number of stages in the causal chain leading to the tokening of a state by a cognizer.  At one extreme, all states of a cognizer happen when only when and because of the big bang.  At the other extreme, S_c will covary with some set of patterns of retinal, LGN, or even striate cortex activation.  Covariationists must find a way to rule out all links in the causal chain except the object, property, event, or relation in the immediate distal environment.  For example, retinal stimulation patterns might be ruled out as candidates for content because of the multiplicity of patterns that cause a single state in later stages of visual processing.

Illustrative Animmation

Non-existent Objects

Creating an account for the representational content of states that seem to represent non-existing objects poses yet another challenge for nomic covariation theorists.  If people have states that represent the property of being a unicorn, it would seem that those states must covary with unicorns.  Given the current paucity of unicorns with which a candidate state can covary, nomic covariation theories do not seem to have a natural way to account for such representational abilities.  Fodor (1990)  claims that his asymmetric dependence theory can handle such cases.  Another possible strategy for the covariationist would be to define non-existing objects, properties, events, or relations using lower-level perceptual features.

Absent Qualia

Some philosophers, for example, Searle (1980), Maudlin (1989), and BonJour (1991) also argue against covariation theories (in fact, against all causal theories) by arguing that cases where, from the internal perspective of an agent, the agent lacks the necessary conscious awareness of content satisfy the nomic covariationists' theory.  Searle, Maudlin, and BonJour then conclude that, since mental states--hence, the content of mental states--requires consciousness, nomic covariation theories fail to explain mental representation.  For these philosophers mental states that have intentionality (are about something) have a qualitative aspect to them.   
 

Searle (1980), for instance, argues that one can manipulate Chinese symbols having the appropriate causal connections to objects or properties in the world without thereby coming to understand Chinese.

Content Fixing, Error, Univocal Contents and Twin Earths

Environmental differences or changes seem to pose a another serious problem for nomic covariationists.  To wit, suppose one discovers the existence of two Earth-like worlds: On one planet, call it Earth¹, where no water exists, but in which deuterium, D²O, (or some chemically different, but phenomenally similar substance) proves plentiful.   On the other planet, call it Earth², H²O proves plentiful, but no D²O exists.  Bob grows up on Earth¹, where he forms beliefs, etc. about "water" as a result of interacting with D²O.  Bill grows up on Earth², where he develops the exact same set of beliefs, etc. about "water" as Bob, but based upon interacting with H²O.  On any nomic covariationist account the two men must have content-non-identical belief contents when thinking about "water."  Bob's states represent H²O.  Bill's states represent D²O.  However, suppose that Bob travels to Earth², he survives by representing H²O as D²O.  His cognitive success seems to result from his systematic misrepresentation of H²O as D²O. 

One response holds that Bob has always represented the property one might describe as H²O or D²O using "water."  Another response would have Bob's representational content change--all at once or over time--upon reaching Earth².  Critics assert that nomic covariationists lack the resources to disambiguate such scenarios.  While most philosophers consider the former even less intuitively plausible than the later, neither response seems to jibe with Bob's conscious experiences.

He seems to exploit the ability to represent D²O (and not H²O) on Earth¹ and he exploits the ability to represent H²O and not D²O) on Earth².  Yet, from a subjective perspective, Bob he never seems to change representational content or undergo any conceptual revolution in order to exploit these representational abilities. 

Functional Role Semantics

Whereas covariationists focus upon a single causal connection in fixing content, advocates of functional role semantics (Block 1986 and 1987, Field 1977 and 1978, Harman 1987), suggest that the overall network of causal relations into which a state can enter fixes its content.  Theorists most often specify causal roles under the functional/computational description of those roles.  Versions of functional role semantics include "Conceptual Role Semantics," "Procedural Semantics," or "Inferential Role Semantics."  Two versions of this approach emerge in the literature, designated "wide" or "long-armed" and "narrow" or "short-armed".  Narrow functional role theorists limit the causal relations that determine content to those occurring between mental states.  Wide theorists allow connections to the distal environment and even social contexts to delimit the content of a state.

Illustrative Animation

Advantages of Functional Role Semantics

Unified Account of Content

Functional role theories have several attractive features.  For instance, functional role theorists do not need bifurcated accounts of representational content.  Recall that the law-like connection between state and property required by covariation theories seems the most plausible for lower-level properties like color or shape.  Therefore, covariationists have difficulties accounting for a state's ability to represent higher-level properties and uninstantiated properties.  Thus, theorists often offer alternative accounts of representational content for these higher-level properties (like definition through lower-level properties).  Functional role semanticists hypothesize that all states get their content in the same manner, via their functional role.  Hence, functional role theorists avoid difficulties in explaining the representation of higher-level and uninstantiated properties.  They need not depend upon bifurcated accounts of representational properties.

Belief Change

Second, functional role theorists can easily accommodate the observation that changes in beliefs can result in changes in representational content, since changing beliefs will often change the functional roles of states.  For example, one standardly associates polio with fever, headache, stiff back and/or neck, muscle pains and/or tenderness.  If polio affects the central nervous system, then one can also often find paralysis of the essential muscles, i.e., those controlling swallowing, heartbeat, and respiration.  Thus, if one knows some of the common symptoms of polio, but lacks modern diagnostic techniques and equipment, one might well suppose that someone has polio when that person actually has one of a number different diseases and conditions having symptoms similar to polio: bites from a snakes with neurotoxic venom, poisoning, diphtheria, myasthenia gravis, Guillain-Barre syndrome, transverse myelitis, tick paralysis, rabies, and even botulism.  But, if one comes to know about the physiological basis of each of these diseases and conditions, their typical causes, and the course of each disease or condition, one can often develop more sophisticated differential diagnoses that separate polio cases from other cases.  Of course, if comes to have the beliefs and instruments involved in contemporary diagnostic techniques and equipment, one's functional roles become even more specific and differential.  But, during such changes of beliefs and other functional roles, does one change the content of one's polio representation?  People often intuitively suppose the content remains fixed despite these changes--but how?

Rationality Constraints on Belief Ascriptions

Finally, sometimes other people's beliefs seem irrational or vacuous from our own perspective or even from the professed beliefs and related functional roles of the person.  For example, the idea of the holy trinity seems self-contradictory to many atheists.  Atheists will ask how God can be one entity while also being three entities.  The logical positivists, in fact, thought it important to declare the vacuity of such beliefs.  Nevertheless, these beliefs seem real psychological entities with real causes and real effect, and especially with real contents.  But, the functional role theorist seems prima facie at a loss to attribute content to such beliefs given that the functional role for such odd-seeming or irrational-seeming beliefs picks out either nothing or no univocal class of things. 

Still, functional role theorists can capture the intuition that such odd-seeming or irrational-seeming belief have content through adopting a principle of charity in content ascription.  Functional role semanticists can argue that in trying to understand the representational content of another cognizer’s state, one is constrained by the heuristic that the overall set of content ascriptions must "make sense."  That is, content ascriptions guided by the principle of charity must suppose that the cognizer's belief system prove consistent with the supposition that overall, the cognizer’s interactions with the world are intelligent or rational.  For example, if someone saw an actual UFO the function role theorist might well suppose that UFO was the content of their representational state.  However, if the person was the head of the UFO-Sceptics association and had a theory debunking the UFO interpretation of their experience.  The functional role theorist would likely not attribute the content "UFO" to the person's state.

Problems for Functional Role Theories
Semantic Holism
Theorists typically raise three distinct but related objections to all versions of functional role semantics.   First, to identify particular states in disparate individuals as states having the same representational content, functional role semantics prima facie requires that the states have identical functional roles.  Fodor (1992, 1987) among others has termed this view Holism, about content (also Semantic Holism).  For example, suppose that it appears that both Bill and Bob believe that "Wallis’ article is enlightening."  That is, they both point to the same article, proclaim that it is enlightening, assign it in their class, etc..  However, the causal roles of their respective states differ in a single respect; Bill believes "Wallis is a pompous know-it-all," while Bob believes "Wallis is a precocious windbag."  Prima facie, functional role semantics dictates that Bill’s belief that "Wallis’ article is enlightening," has a different content than Bob’s.  Similarly, people with disparate cognitive or perceptual abilities seem to have disparate representational contents.  Thus, the theory appears to imply that the average human has a different concept of traffic light than color blind humans.

Functional role semanticists generally adopt one of two responses to the above objection.  Both responses constitute variations of the two-classes strategy.  One response strategy would prune the number and/or kind of causal or computational connections necessary for belief/content identity, thereby allowing for belief/content identity across individuals with somewhat different causal/computational roles.  In the above case, differences in beliefs that are peripheral to the belief that "Wallis’ article is enlightening" would not necessarily constitute the basis for content non-identity.   In effect, one creates a content-imbuing set of cases in which the content-imbuing functional roles operate exclusively.  One likewise creates a content-utilization set of cases in which all the individuals functional roles operate to generate tokens of states; some tokenings resulting from the wider set of functional roles can prove non-veridical without altering the content of the state.  The overall rationale for such a move consists in the supposition that a core set of essential or content-imbuing functional roles exists for any object, property, event, or relation.  Thus, one can differ with regard to non-essential or content-utilizing functional roles so long as one has the essential or content-imbuing roles.

While the just-rehearsed response has intuitive appeal, critics (Fodor 1988) point to the potential difficulties in distinguishing core (central) beliefs or other causal links from peripheral ones in a manner that is not hopelessly unsystematic and ad hoc.  For instance, what causal connections (beliefs, desires, dispositions to take action, etc..) constitute the core of one’s belief that the colorless, tasteless, odorless liquid before one is water.  Need one know, to take a case, of the existence of deuterium oxide (heavy water), make appropriate inferences with regard to D²O, discriminate between D²O and H²O, etc..  One common suggestion for distinguishing core, content-imbuing beliefs etc., from peripheral beliefs etc. asserts that content-imbuing beliefs etc. include only analytic beliefs etc. in the set of core beliefs etc..  Analytic truths are conceptual truths, and as such prove true of something solely in virtue of the nature of the thing/concept.  Such a suggestion requires a real distinction between analytic beliefs and non-analytic (synthetic) beliefs.  Many philosophers believe that Quine (1953/1970) has effectively undermined the robustness of such a distinction.  Similarly, Stich (1983) argues that judgments of belief/content identity lack the intuitive precision that advocates of the analytic distinction would seem predict.

What about second response to the prima facie difficulty of intuitively identical content across differing functional roles?  The second response claims that while such beliefs lack content-identity, yet such beliefs prove strongly content-similar.  The key move in this response involves denying the necessity of strict content-identity (i.e., that content-identity is not a binary notion), and adopting a notion of content-similarity that ranges from completely non-identical to completely identical (Cummins 1989).  The strength (or weakness) of this response lies in its ability to accommodate the intuition that Bill’s and Bob’s beliefs have the same content while simultaneously acknowledging the theoretical constraint that differences in functional roles dictate differences in content.

Critics of functional role semantics (ex. Fodor, 1988) argue that a graded notion of content identity, in addition to being counterintuitive, undermines the ability to formulate psychological generalizations and subsume particular cases under those generalizations.  Content-similarity reduces cognitive science, these critics claim,  to the unworkable notion that Bill’s and Bob’s, say, 97% content similar Wallis beliefs can provide adequate explanations of their cognitive behaviors.  Say, the beliefs are 97% subsumed under generalizations regarding the Wallis belief.  Furthermore, Fodor (1992 and 2001) claims colloquial notions of belief similarity like, "His notion of mental representation is similar to mine," as well as their theoretical counterparts presuppose a notion of content-identity--a notion such holistic views, Fodor Claims, cannot provide in any cases where beliefs diverge.

Content Fixing, Error, Univocal Contents and Twin Earths
A closely related objection to that of holism is the argument that functional role theories have no non-arbitrary way of fixing content. Specifically, functional role semanticists either (1) cannot distinguish representational error from veridical representation, (2) their theory results in non-univocal content ascriptions, and/or (3) their theory results in multiple content ascriptions applying simultaneously to a given state.  For example, suppose one discovers the existence of two Earth-like worlds: one, call it Earth¹  in which there is no water, but in which D²O (or some chemically different, but phenomenally similar substance) is plentiful; the other, call it Earth², where H²O is plentiful, but there is no D²O.  Bob grows up on Earth¹, where he forms beliefs, etc. about "water" as a result of interacting with D²O.  Bill grows up on Earth², where he develops the exact same set of beliefs, etc. about "water" as Bob, but based upon interacting with H ²O.  If one is a narrow functional role theorist (using only causal connections "internal" to the cognizer), then the two men must have identical belief content when thinking about "water".  They either form beliefs, make inferences, etc., using states that represent what one would describe using the disjunction, "H²O or D²O", or their states represent both H²O and also D²O (as distinct entities) simultaneously.  Most philosophers consider the latter even less intuitively plausible than the former.

Long arm or two factor theories can distinguish Bill’s and Bob’s beliefs since they include causal connections to environmental objects and/or properties.  However, suppose that Bob also has the beliefs that "water is H²O," that "I live on Earth²," and "Earth² has only H²O on its surface."    Is Bob representing "water" as  D²O, but forming a false belief about the chemical structure of D²O?  Or, is he representing "water" as H²O and forming false beliefs about the D²O he finds in his environment.  Indeed, suppose one moves Bob to Earth²--do his contents now become totally veridical?   
  
Critics assert that functional role semantics lacks the resources to disambiguate such scenarios.  While most philosophers consider the former multiple content scenario even less intuitively plausible than the latter disjunctive content scenario, neither response seems to jibe with Bob's conscious experiences.  He seems to exploit the ability to represent D²O (and not H²O) on Earth¹ and he exploits the ability to represent H²O (and not D²O) on Earth².  Yet, from a subjective perspective, Bob never seems to change representational content or undergo any conceptual revolution in order to exploit these representational abilities. 

Representational Structure
Finally,  Fodor and LePore (1992), Fodor and McLaughlin (1981), and Fodor and Plysyhyn (1988) raise the objection that functional role semantics seems to violate truths about the structure of language and thought; specifically compositionality and systematicity.  Compositionality is the theory that the meaning of a complex expression in a language results from the meanings of its constitutive elements.   Compositionality plays an central role in many linguistic theories.  Supposing compositionality's truth for both language and thought provides theorists with a fairly straightforward explanation of the human ability to grasp an enormous number of different thoughts of varying complexity and their corresponding linguistic expressions.  For instance, because one understands the individual elements of a sentence, like "cup" and "coffee," one understands the complex expressions built from those simple elements, like "cup of coffee," "hot coffee," "coffee gives me the shakes," etc..  One explains one's understanding by noting that the meaning of these complex sentences is built-up from the meaning of their constitutive elements.

Fodor and others claim that the functional role of a complex, non-idiomatic representation fails result at times from a function of the functional roles of its parts. As a result, functional role theories can represent "cup of coffee" without having the concepts and associated inferences of "cup" or "coffee".  This alleged aspect of functional role representation schemes would result in the possibility that one could represent and think about "cup of coffee," but could not represent or think about "iced coffee," "hot coffee," etc..  Nor could one understand or evaluate the sentence, "Coffee is the legal drug for which we exploit the people and natural resources of many third world countries."

Block (forthcoming ) suggests that functional role semanticists can account for compositionality by adopting context-sensitive rules of use for words.  That is, theorists can understand meanings of words as sets of inferences (functional roles) from words to sentences containing the word and sets of inferences from sentences to containing the words.  Thus, "cup of coffee" can get its meaning from cup in a manner consistent with one's understanding "iced coffee", etc..  Connectionists with commitments to functional roles and others (Smolensky 1991 and 1991a) argue that connectionist systems can have constituents satisfying many of the criteria of compositionality without adopting traditional compositional representational frameworks.

Causal Theories and Cognitive Science
Cognitive science proceeds in the absence of a resolution of the debate over theories of representation.  However, the adoption of either nomic covariance theories or functional role semantics does shape the research of individual cognitive scientists.  Many important debates in cognitive science result from allegiance to one or the other causal theory.  For example, one aspect of the connectionist vs Turing-compatible approaches debate is that advocates of connectionist frameworks tend to adopt functional role semantics, while advocates of Turing-compatible frameworks tend to adopt nomic covariation.  Similarly, theorists draw the debate over the nature of mental imagery along covariance vs functional role lines.  Indeed, many people suppose that the adoption of one or the other theory of mental content by the majority of cognitive scientists will have a profound impact on the field.
 

Bibliography

• Akins, K. (1996) "Of Sensory Systems and the 'Aboutness' of Mental States." Journal of Philosophy. 93, 337-372.
• Allen, C.  (forthcoming) "A Tale of Two Froggies." Canadian Journal of Philosophy.
• Almog, J., Perry, J. and Wettstein, H., eds. Themes from Kaplan , New York: Oxford University Press (1989).
• Anquetil, T. and M. Jeannerod, "Simulated Actions in the First and in the Third Person Perspectives Share Common Representations." Brain Research, 2007. 1130: p. 125-129.
• Aristotle. De Anima, in The Complete Works of Aristotle: The Revised Oxford Translation, Oxford: Oxford University Press (1984).
• Antony, L. & Levine, J. " The nomic and the robust," In (B. Loewer & G. Rey, eds) Meaning in Mind: Fodor and his Critics. Blackwell (1991).
• Barsalou, L.W., "The Instability of Graded Structure: Implications for the Nature of Concepts," in Concepts and Conceptual Development: Ecological and Intellectual Factors in Categorization, U. Neisser, Editor. 1987, Cambridge University Press: Cambridge, UK.
• Barsalou, L.W. and D.L. Medin, "Concepts: Static Definitions or Context-Dependent Representations?" Cahiers de Psychologie, 1986. 6(2): p. 187-202.
• Beaney, M., ed. The Frege Reader, Oxford: Blackwell Publishers (1997).
• Berkeley, G. Principles of Human Knowledge, in M.R. Ayers, ed., Berkeley: Philosophical Writings, London: Dent (1975).
• Biederman, I., "Recognition-by-components: A Theory of Human Image Understanding". Psychological Review, 1987. 94(2): p. 115-117.
• Block, N. (ed.). Readings in Philosophy of Psychology, Vol. 2 , Cambridge, Mass.: Harvard University Press (1981).
• Block, N. (ed.) Imagery, Cambridge, Mass.: The MIT Press (1982).
• Block, N. "Mental Pictures and Cognitive Science," Philosophical Review 93 (1983): 499-542.
• Block, N. "Advertisement for a Semantics for Psychology," in P.A. French, T.E. Uehling and H.K. Wettstein, eds., Midwest Studies in Philosophy, Vol. X, Minneapolis: University of Minnesota Press (1986): 615-678.
• Block, N.  "Functional Role and Truth Conditions," Proceedings of the Aristotelian Society suppl.  (1987) 61: 157-181.
• Block, N. "Mental Paint and Mental Latex," in E. Villanueva, ed., Philosophical Issues, 7: Perception (1996): 19-49.
• Boghossian, P.A. "Content," in J. Kim and E. Sosa, eds., A Companion to Metaphysics, Oxford: Blackwell Publishers Ltd. (1995): 94-96.
• BonJour, L. "Is Thought a Symbolic Process?" Synthese 89 (1991) 331-352.
• Burge, T. "Individualism and the Mental," in P.A. French, T.E. Uehling and H.K. Wettstein, eds., Midwest Studies in Philosophy, Vol. IV, Minneapolis: University of Minnesota Press (1986): 73-121.
• Chalmers, D. The Conscious Mind, New York: Oxford University Press (1996).
• Chalmers, D. "Connectionism and Compositionality: Why Fodor and Pylyshyn Were Wrong," Philosophical Psychology 6 (1993): 305-319.
• Chisholm, R. and Sellars, W. "The Chisholm-Sellars Correspondence on Intentionality," in H. Feigl, M. Scriven and G. Maxwell, eds., Minnesota Studies in the Philosophy of Science, Vol. II, Minneapolis : University of Minnesota Press (1958): 529-539.
• Chomsky, N. Aspects of the Theory of Syntax, Cambridge, Mass.: The MIT Press (1965).
• Churchland, P.M. "Eliminative Materialism and the Propositional Attitudes," Journal of Philosophy 78 (1981): 67-90.
• Churchland, P.M. "On the Nature of Theories: A Neurocomputational Perspective," in W. Savage, ed., Scientific Theories: Minnesota Studies in the Philosophy of Science, Vol. 14, Minneapolis: University of Minnesota Press (1989): 59-101.
•  Churchland, P and Sejnowski, T (1992) The Computational Brain. Cambridge, MA: MIT Press.
• Cummins, R (1989) Meaning and Mental Representation.  Cambridge: MIT Press.
• Cummins, R. (1996) Representations, Targets, and Attitudes.  Cambridge: MIT Press.
• Davidson, D. "Radical Interpretation," Dialectica 27 (1973): 313-328.
• Davidson, D. "Belief and the Basis of Meaning," Synthese 27 (1974): 309-323.
• Davidson, D. "Thought and Talk," in S. Guttenplan, ed., Mind and Language, Oxford: Clarendon Press (1975): 7-23.
• Davidson, D. "Rational Animals," Dialectica 4 (1982): 317-327.
• Davidson, D. "Knowing One's Own Mind", Proceedings of the American Philosophical Association (1987).
• Dennett, D. Content and Consciousness, London: Routledge and Kegan Paul (1969).
• Dennett, D. "The Nature of Images and the Introspective Trap," pages 132-141 of Dennett 1969, reprinted in Block 1981 (1981): 128-134.
• Dennett, D. The Intentional Stance, Cambridge, Mass.: The MIT Press (1987).
• Dennett, D. "True Believers: The Intentional Strategy and Why it Works," in Dennett 1987 (1987a): 13-35.
• Dennett, D. "Reflections: Real Patterns, Deeper Facts, and Empty Questions," in Dennett 1987 (1987b): 37-42.
• Dennett, D. "Quining Qualia," in A.J. Marcel and E. Bisiach, eds., Consciousness in Contemporary Science, Oxford: Clarendon Press (1988): 42-77.
• Dennett, D. "Real Patterns," The Journal of Philosophy LXXXVII (1991): 27-51.
• Devitt, M. Coming to Our Senses: A Naturalistic Program for Semantic Localism, Cambridge: Cambridge University Press (1996).
• Dretske, F. Seeing and Knowing, Chicago: The University of Chicago Press (1969).
• Dretske, F. Knowledge and the Flow of Information, Cambridge, Mass.: The MIT Press (1981).
• Dretske, F. Explaining Behavior: Reasons in a World of Causes , Cambridge, Mass.: The MIT Press (1988).
• Dretske, F. "The Need to Know," in  Lehrer, K. and Clary, M. eds., Theory of Knowledge: The State of the Art Tucson, AZ: University of Arizona Press (1989).
• Dretske, F. Naturalizing the Mind, Cambridge, Mass.: The MIT Press (1994).
• Evans, G. The Varieties of Reference, Oxford: Oxford University Press (1982).
• Field, H.  "Logic, Meaning, and Conceptual Role," Journal of Philosophy  (1977) 74: 379-409.
• Field, H. "Mental representation," Erkenntnis 13 (1978): 9-61.
• Flanagan, O. Consciousness Reconsidered, Cambridge, Mass.: The MIT Press (1992).
• Fodor, J.A. The Language of Thought, Cambridge, Mass.: Harvard University Press (1975).
• Fodor, J.A. "Propositional Attitudes," The Monist 61 (1978): 501-523.
• Fodor, J.A. Representations, Cambridge, Mass.: The MIT Press (1981).
• Fodor, J.A. "Introduction," in Fodor 1981 (1981a): 1-31.
• Fodor, J.A. "Methodological Solipsism Considered as a Research Strategy in Cognitive Psychology," in Fodor 1981 (1981b): 225-253.
• Fodor, J.A. "The Present Status of the Innateness Controversy," in Fodor 1981 (1981c): 257-316.
• Fodor, J.A. "Cognitive Science and the Twin-Earth Problem," Notre Dame Journal of Formal Logic 23 (1982): 98-118.
• Fodor, J.A. Psychosemantics, Cambridge, Mass.: The MIT Press (1987).
• Fodor, J. and Pylyshyn, Z. (1988). "Connectionism and Cognitive Architecture: A Critical Analysis," in Cognition. 28: 3-71.
• Fodor, J.A. A Theory of Content and Other Essays, Cambridge, Mass.: The MIT Press (1990a).
• Fodor, J.A. "Psychosemantics or: Where Do Truth Conditions Come From?" in W.G. Lycan, ed., Mind and Cognition: A Reader, Oxford: Blackwell Publishers (1990), (1990b): 312-337.
• Fodor, J. and McLaughlin, B. (1991). "Connectionism and the Problem of Systematicity: Why Smolensky's Solution Doesn't Work," in T. Horgan and J. Tienson (eds.) Connectionism and the Philosophy of Mind. Boston: Kluwer Academic Publishers.
• Fodor, J. and LePore, E.  Holism: A Shoppers’ Guide. Oxford: Blackwell. (1992).
• Fodor, J.A. The Elm and the Expert, Cambridge, Mass.: The MIT Press (1994).
• Fodor, J. The Mind Doesn’t Work That Way.  Cambridge: MIT Press.  (2001).
• Fodor, J.A. and Pylyshyn, Z. "How Direct is Visual Perception?: Some Reflections on Gibson?s ?Ecological Approach?,"Cognition 9 (1981): 207-246.
• Fodor, J.A. and Pylyshyn, Z. "Connectionism and Cognitive Architecture: A Critical Analysis," Cognition 28 (1988): 3-71.
• Frege, G. The Foundations of Arithmetic, trans. J.L. Austin, New York: Philosophical Library (1953).
• Frege, G. "On Sinn and Bedeutung" (1892) in Beany 1997: 151-171.
• Frege, G. "Thought" (1918) in Beany 1997: 325-345.
• Geach, P. Mental Acts: Their Content and Their Objects, London: Routledge & Kegan Paul (1957) .
• Goldman, A. "The Psychology of Folk Psychology," Behavioral and Brian Sciences 16 (1993): 15-28.
• Goodman, N. Languages of Art (2nd ed.), Indianapolis: Hackett (1976).
• Grice, H.P. "Meaning," Philosophical Review, 66 (1957): 377-388; reprinted in Studies in the Way of Words, Cambridge, Mass.: Harvard University Press (1989): 213-223.
• Harman, G. Thought, Princeton: Princeton University Press (1973).
• Harman, G. "(Non-Solipsistic) Conceptual Role Semantics," in E. Lepore, ed., New Directions in Semantics, London: Academic Press (1987): 55-81.
• Harman, G. "The Intrinsic Quality of Experience," in J. Tomberlin, ed., Philosophical Perspectives 4: Action Theory and Philosophy of Mind , Atascadero: Ridgeview Publishing Company (1990): 31-52.
• Haugeland, J. "Analog and analog," Philosophical Topics 12 (1981): 213-226.
• Heil, J. "Being Indiscrete," in J. Greenwood, ed., The Future of Folk Psychology, Cambridge: Cambridge University Press (1991): 120-134.
• Horgan, T. and Tienson, J. Connectionism and the Philosophy of Psychology, Cambridge, Mass: The MIT Press (1996).
• Horst, S. Symbols, Computation, and Intentionality, Berkeley: University of California Press (1996).
• Hubel, D. and Wiesel, T.  (1977) "Ferrier Lecture: Functional Architecture of Macaque Monkey Visual Cortex."  Proc. R. Soc. Lond.  (1977) 198:1-59.
• Hume, D. A Treatise of Human Nature, L.A. Selby-Bigge, ed., revised P.H. Nidditch, Oxford: Oxford University Press (1978).
• Jackendoff, R. Computation and Cognition, Cambridge, Mass.: The MIT Press (1987).
• Johnson-Laird, P.N. Mental Models, Cambridge, Mass.: Harvard University Press (1983).
• Johnson-Laird, P.N. and Wason, P.C. Thinking: Readings in Cognitive Science, Cambridge University Press (1977).
• Kaplan, D. "Demonstratives," in Almog, Perry and Wettstein 1989 (1989): 481-614.
• Kosslyn, S.M. Image and Mind, Cambridge, Mass.: Harvard University Press (1980).
• Kosslyn, S.M. "The Medium and the Message in Mental Imagery," in Block 1982 (1982): 207-246.
• Kosslyn, S. Ghosts in the Mind?s Machine, New York: W.W. Norton & Co. (1983).
• Kosslyn, S.M. and Pomerantz, J.R. "Imagery, Propositions, and the Form of Internal Representations," Cognitive Psychology 9 (1977): 52-76.
• Leeds, S. "Qualia, Awareness, Sellars," Noûs XXVII (1993): 303-329.
• Lerdahl, F. and Jackendoff, R. A Generative Theory of Tonal Music , Cambridge, Mass.: The MIT Press (1983).
• Levine, J. "On Leaving Out What It?s Like," in M. Davies and G. Humphreys, eds., Consciousness, Oxford: Blackwell Publishers (1993): 121-136.
• Levine, J. "On What It Is Like to Grasp a Concept," in E. Villanueva, ed., Philosophical Issues 6: Content, Atascadero: Ridgeview Publishing Company (1995): 38-43.
• Lewis, D. "Analog and Digital," Noûs 5 (1971): 321-328.
• Lewis, D. "Radical Interpretation," Synthese 23 (1974): 331-344. (Reprinted, with Postscript, in Lewis 1983: 108-121.)
• Lewis, D. Philosophical Papers, Vol. I, New York: Oxford University Press (1983).
• Loar, B. Mind and Meaning, Cambridge: Cambridge University Press (1981).
• Locke, J. An Essay Concerning Human Understanding, P.H. Nidditch, ed., Oxford: Oxford University Press (1978).
• MacDonald, C. and MacDonald, G. Connectionism: Debates on Psychological Explanation, Oxford: Blackwell Publishers (1995).
• Maddox, et al., "Category Label and Response Location Shifts in Category Learning." Psychological Research, 2010. 74(2): p. 219-236.
• Malt, B. and Smith, E., "Correlated Properties in Natural Categories." Journal of Verbal Learning & Verbal Behavior, 1984. 23(2): p. 250-269.
• Marr, D. Vision, New York: W.H. Freeman and Company (1982).
• Martin, C.B. "Proto-Language," Australasian Journal of Philosophy 65 (1987): 277-289.
• Maudlin, T.  "Computation and Consciousness," Journal of Philosophy 86 (1989) 407-432.
• McCulloch, W.S. and Pitts, W. "A Logical Calculus of the Ideas Immanent in Nervous Activity," Bulletin of Mathematical Biophysics 5 (1943): 115-33.
• McGinn, C. "The Structure of Content," in A. Woodfield, ed., Thought and Content, Oxford: Oxford University Press (1982): 207-258.
• McGinn, C. Mental Content, Oxford: Blackwell Publishers (1989).
• McGinn, C. "Content and Consciousness," in C. McGinn, The Problem of Consciousness, Oxford: Blackwell Publishers (1992): 23-43.
• Medin, D. and Shoben, E., "Context and Structure in Conceptual Combination." Cognitive Psychology, 1988. 20(2): p. 158-190.
• Millikan, R. Language, Thought and other Biological Categories , Cambridge, Mass.: The MIT Press (1984).
• Millikan, R.  "Thought Without Laws: Cognitive Science Without Content." Philosophical Review (1986) 95:47-80.
• Neander, K. "Misrepresenting and Malfunctioning." Philosophical Studies 79:109-41 (1995).
• Neander, K. "Dretske's Innate Modesty," in  Australasian Journal of Philosophy 74:258-74 (1996).
• Newell, A. and Simon, H.A. Human Problem Solving, New York: Prentice-Hall (1972).
• Newell, A. and Simon, H.A. "Computer Science as Empirical Inquiry: Symbols and Search," Communications of the Association for Computing Machinery 19 (1976): 113-126.
• Osherson, D.N., Kosslyn, S.M. and Hollerbach, J.M. Visual Cognition and Action: An Invitation to Cognitive Science, Vol. 2,   Cambridge, Mass.: The MIT Press (1990).
• Papineau, D. "Representation and Explanation," in Philosophy of Science 51:550-72 (1984).
• Papineau, D. Reality and Representation, Oxford: Blackwell Publishers (1987).
• Peacocke, C. Sense and Content, Oxford: Clarendon Press (1983).
• Peacocke, C. "Perceptual Content," in Almog, Perry and Wettstein 1989 (1989): 297-329.
• Peacocke, C. "Scenarios, Concepts and Perception," in T. Crane, ed., The Contents of Experience, Cambridge: Cambridge University Press (1992): 105-35.
• Pinker, S. Learnability and Cognition, Cambridge, Mass.: The MIT Press (1989).
• Port, R. and Van Gelder, T. Mind as Motion: Explorations in the Dynamics of Cognition, Cambridge, Mass.: The MIT Press (1995).
• Putnam, H. "The Meaning of Meaning," in Philosophical Papers, Vol. 2, Cambridge: Cambridge University Press (1975): 215-71.
• Pylyshyn, Z. "The Rate of ?Mental Rotation? of Images: A Test of a Holistic Analogue Hypothesis," Memory and Cognition, 7 (1979): 19-28.
• Pylyshyn, Z. "Imagery and Artificial Intelligence," in Block 1981 (1981a): 170-194.
• Pylyshyn, Z. "The Imagery Debate: Analog Media versus Tacit Knowledge," Psychological Review 88 (1981b): 16-45.
• Pylyshyn, Z. Computation and Cognition, Cambridge, Mass.: The MIT Press (1984).
• Quine, W. "Two Dogmas of Empricism" in Philosophical Review  60: 20-43 (1951).
• Quine, W.  From a Logical Point of View. Cambridge: Harvard University Press.   (1953).
• Quine, W. Word and Object. Cambridge: MIT Press.  (1960).
• Raffman, D. "The Persistence of Phenomenology," in T. Metzinger, ed., Conscious Experience, Paderborn: Schöningh/Imprint Academic (1995): 293-308.
• Ramsey, W., Stich, S. and Garon, J. "Connectionism, Eliminativism and the Future of Folk Psychology," Philosophical Perspectives 4 (1990): 499-533.
• Rey, G. "Introduction: What Are Mental Images?" in Block(1981): 117-127.
• Rey, G. "Sensations in a Language of Thought," in E. Villaneuva, ed., Philosophical Issues 1: Consciousness, Atascadero: Ridgeview Publishing Company (1991): 73-112.

• Rosch, E., "Cognitive Representations of Semantic Categories." Journal of Experimental Psychology: General, 1975. 104(3): p. 192-233.

• Rosch, E., "Natural Categories." Cognitive Psychology, 1973. 4(3): p. 328-350.

• Rosch, E. and Mervis, C., "Family Resemblances: Studies in the Internal Structure of Categories." Cognitive Psychology, 1975. 7(4): p. 573-605.

• Rumelhart, D.E. "The Architecture of the Mind: A Connectionist Approach," in M.I. Posner, ed., Foundations of Cognitive Science, Cambridge, Mass.: The MIT Press (1989): 133-159.
• Rumelhart, D.E. and McCelland, J.L. Parallel Distributed Processing, Vol. I, Cambridge, Mass.: The MIT Press (1986).
• Schiffer, S. 1987. Remnants of Meaning, Cambridge, Mass.: The MIT Press (1987).
• Schiffer, S. "Introduction to the Paperback Edition," in Meaning , Oxford: Clarendon Press (1988): xi-xxix.
• Searle, J.  "Minds, Brians, and Programs," Brain and Behavioral Sciences 3 (1980) 417-457.
• Searle, J.R. Intentionality, Cambridge: Cambridge University Press (1983).
• Searle, J.R. The Rediscovery of the Mind, Cambridge, Mass.: The MIT Press (1990).
• Sellars, W. "Empiricism and the Philosophy of Mind," in K. Gunderson, ed., Minnesota Studies in the Philosophy of Science, Vol. I, Minneapolis: University of Minnesota Press (1956): 253-329.
• Shapiro, L. 1992. Darwin and disjunction: Foraging theory and univocal assignments of content. Philosophy of Science Association 1992, 1:469-80.
• Shapiro, L. 1996. Representation from bottom to top. Canadian Journal of Philosophy 26:523-42.
• Shapiro, L. "Multiple Realizations," The Journal of Philosophy , vol. 97, no. 12, pp. 635-654, (2000).
• Shepard, R.N. and Cooper, L. Mental Images and their Transformations , Cambridge, Mass.: The MIT Press (1982).
• Shoemaker, S. "Qualities and Qualia: What?s in the Mind?" Philosophy and Phenomenological Research 50 (1990): 109-31.
• Smolensky, P. "On the Proper Treatment of Connectionism," Behavioral and Brain Sciences, 11 (1988): 1-74.
• Smolensky, P. "Connectionist Modeling: Neural Computation/Mental Connections," in L. Nadel, L.A. Cooper, P. Culicover and R.M. Harnish, eds., Neural Connections, Mental Computation Cambridge: The MIT Press (1989): 49-67.
• Smolensky, P.   "Connectionism, Constituency, and the Language of Thought," in G. Rey and B. Loewer (eds.) Meaning in Mind .  Cambridge: Blackwell (1991) : 201-227.
• Smolensky, P.  "The Constituent Structure of Connectionist Mental States: A Reply to Fodor and Pylyshyn," in T. Horgan and J. Tienson (eds.) Connectionism and the Philosophy of Mind.  Boston: Kluwer Academic Publishers. (1991a).
• Sterelny, K. "Fodor's Nativism," in Philosophical Studies 55 (1989): 119-141.
• Sterelny, K. The Representational Theory of Mind: An Introduction.  Cambridge: Blackwell (1990).
• Stich, S. 1978. "Autonomous Psychology and the Belief-Desire Thesis," The Monist 61 (1978): 573-591.
• Stich, S. From Folk Psychology to Cognitive Science, Cambridge, Mass.: The MIT Press (1983).
• Strawson, G. Mental Reality, Cambridge, Mass.: The MIT Press (1994).
• Tye, M. The Imagery Debate, Cambridge, Mass.: The MIT Press (1991).
• Tye, M. Ten Problems of Consciousness, Cambridge, Mass.: The MIT Press (1995).
• Tye, M. Consciousness, Color, and Content, Cambridge, Mass.: The MIT Press (2000).
• Van Gelder, T. "What Might Cognition Be, if not Computation?", Journal of Philosophy XCI (1995): 345-381.
• Von Eckardt, B. 1993. What Is Cognitive Science?, Cambridge, Mass.: The MIT Press (1993).
• Wallis, C. "Ceteris Paribus Laws and Psychological Explanation," Philosophy of Science Association 1994. (1994)1: 388-397.
• Wallis, C.  "Representation and the Imperfect Ideal," Philosophy of Science. (1994a) 61: 407-428.
• Wallis, C.  "Asymmetric Dependence, Representation, and Cognitive Science," Southern Journal of Philosophy. (1995) 33:373-401.

Other Internet Resources

• A Field Guide to the Philosophy of Mind
• Dictionary of Philosophy of Mind
• Internet Encyclopedia of Philosophy
• Mind/Brain Resources
• Block’s Article on Functional Role Semantics

Related Entries

Theories

Nomic Covariation

Advantages of Covariation Theories

Covariation Theories and the Disjunction Problem

Functional Role Semantics

Advantages of Functional Role Semantics

Problems for Functional Role Semantics

Causal Theories and Cognitive Science

Bibliography

Other Internet Resources

Related Articles

Theories

Nomic Covariation

Advantages of Covariation Theories

Covariation Theories and the Disjunction Problem

Functional Role Semantics

Advantages of Functional Role Semantics

Problems for Functional Role Semantics

Causal Theories and Cognitive Science

Bibliography

Other Internet Resources

Related Articles