Abstracts of
"Intentionality: Past and Present"

Miskolc, Hungary, June 21-23 2002.

(In alphabetical order. For programme and speakers list, with links: http://hps.elte.hu/intentionality/Programme.html )

Laird Addis

The University of Iowa

Gergely Ambrus

Dept. of Phil.; University of Miskolc

Philip J. Bartok

Department of Philosophy Uiversity of Notre Dame

Ned Block

New York University New York USA

Filip Buekens

Tilburg University, The Netherlands

According to Tye (1996), and a respectful minority in philosophy of mind (Anscombe 1965, Harman 1990, Dretske 1995), qualia are intentional states. They non-conceptually represent properties of objects, events or states in the world. McGinn (1989) and Block (1995), among others, reject this view. I will argue that while the non-representationalist, defending the intrinsic, non-relational character of the 'what it is like to'-property (the feel of the experience) is correct, we still have to explain (or explain away) the representationalist intuition. I will defend that the alleged representational features of experiences derive from our tendency to conceptualize experiences in terms of their (normal) causes and or effects. This account (a) explains why the context 'what it is like to ____' seems intensional, (b) under which conditions we can cognitively derive from knowledge that we have an experience knowledge about its cause , and (c) shows that experiential concepts have two components: a phenomenal component and a causal component. My argument will focus on Tye's defense of the intentional character of experiences.

The argument runs as follows:

It is constitutive for an experience E that it manifests itself as E to a subject. (This captures the Kripkean intuition that, say, pain is essentially the feeling of pain.)

The constitutive principle requires a de re reading: the intuition cannot assume that the subject of E has the corresponding E-concept. (A suitably re-wired Martian can undergo pain without grasping the concept of pain, and a subject can grasp the non-phenomenal component of an experiential concept E without ever having experienced E)

Filling in the phenomenal component of an E-concept requires recognitional knowledge of that experience (Loar 1990, Perry 2001). This requires propositional knowledge that this [the experience] is what it is like to undergo E. This kind of knowledge is conceptual with respect to the concept E, and essentially indexical with respect to the experience the subject undergoes.

The alleged intensionality of the context 'what it is like to ______' (Tye) derives from its being embedded in the broader, unproblematically intensional context 'X knows that this is what it is like to _____'

In 'X knows that this is what it is like to have E', "E" (the concept) can be substituted by redescriptions of that experience in terms of its cause: if X knows that this is what it is like to have pain, and X knows that this experience was caused by being hit by a stone, then X knows knows that this is what it is like to have an experience caused by being hit by a stone. The experience itself, the feel, does not represent its cause, but given that the subject knows its cause (or has beliefs about it), she can redescribe that experience in terms of its cause.

Most of our E-concepts have the causal component already built-in in the concept itself: 'knowing what it is like to taste Tokay', knowing what it is like to see red' (red being the color one perceives), 'knowing what it is like to be weightless', etc.

The causal nature of many of our E-concepts explain how we can manage to avoid certain experiences (avoiding their normal causes) and tell us what to do if we desire to enjoy them.

Conclusion: an experience is a local and an intrinsic phenomenon. E-concepts have not only a phenomenal component (when the concept is recognitional), but also a causal component (giving us information about the normal cause of the experience). The latter component cannot be reduced to the former (experiences don't represent), and the former cannot be reduced to the latter (the qualitative character of an experience cannot be functionalized away). Experiences are not intentional states.

Arkadiusz Chrudzimski

University of Salzburg & University of Zielona Gora

Tim Crane

University College London

Sean Crawford

Philosophy Department, the Open University, UK

István Danka

Hungarian Academy of Sciences Institute for Philosophy

In my present paper I will discuss the question of how Donald Davidson’s ‘anomalous monism’ saves intentionality, and gives a causalist account of human actions. If I understood Davidson right, then his theory of descriptions give a new explanation of act individuation, which goes over the ‘Unifier-Multiplier Controversy’, as Karl Pfeifer called the problem of whether (e.g.) killing someone is an event or a chain of events. I will answer Goldman’s arguments against Davidsonian act-individuation, and point out, that Davidson wants to transform the multiplicity of events to multiplicity of ways of descriptions of the same event.

Explainig the intentionality is a good place to argue for Davidson is not a unifier. I will take more and more complex examples, and show that Goldman’s main objection against Davidson, i.e. Davidson would think that causality is depend on the language, is right in case of Goldman himself, but not in case of Davidson.

Anscombe, and Davidson at first sight, think that killing someone is one and only one action, which has many descriptions as ‘moving his/her finger’, ‘pulling the trigger’, ‘cause someone’s death’, etc. These are not different actions but different descriptions for the same action.

Goldman argues against this view in his book Theory of Human Action. He claims that we should abandon of the unity of a killing for bypassing contradictions involved the indentification of events in so-called by-relation. By-relation, according to Goldman, is irreflexive and asymmetric, however, as we know, identity is reflexive and symmetric, hence it is not possible that events e, e’, by -related to one another, could be identical. If Goldman is right, then there are some actions without causal relation, but by-related to each other. And, according to Davidson, every event could have only two ways of relating to others, namely causally or identically. (Goldman have two other arguments against the Davidsonian view, as Pfeifer shows, but I think that the first is only a special case of the above-mentioned relational objection, still the second, ‘temporal’ claim is a mistake, because temporally distinct events are not same in Davidsonian view at first sight, too.)

After Goldman, as he thinks, destroys the Davidsonian theory, he works out his idea of level generation, which is grounded on the concepts of ‘act token’ and ‘act type’. I will point out, that these categories involve a kind of Platonism, for it is a necessary condition for applying them, that universal entities must have epistemological, and, maybe, ontological priority over particulars. I think that the only way to evade this kind of metaphysical realism is to say that our act types are just theoretical aids for explaining act-individuation. Thus we come to Davidson’s transcendentalist point, that events can’t be considered in isolation from their descriptions. Hence for we save the idea that causality is a physical, and not a linguistic category, we must suppose, that there is something outside our descriptions, and that kind of reality can’t be analysed but just presupposed. What is analysable, is an event under descriptions.

But how the attack against Davidson could be solved? Monroe C. Beardsley has pointed out, that the term ‘by -relation’ has an intentional connotation. Saying that ‘doing something by doing sometng else’ presupposes an agent, who can do it. And, if I am not mistaken, this is the point, which can help us to understand that Davidson is not a theorist of the Unifier-model. Since I think that Davidson doesn’t say that to kill someone is one and only one event independent of descriptions , but he says that it is one and only one event under some descriptions. Davidson thinks that sometimes we describe events in terms of their causes or effects. When we take an event under two different descriptions as by-related, either of them is a description in terms of (some) causes or effects of the event. According to Davidson, all events can be described in physical terms, i.e. in terms of causality - this is the reason for accepting that the world is physical rather than mental. But some events can be described in terms of mentalism - these events are the so-called actions: events understood as intentional ones. There’s no intentionality in the world - or, if there would be, we couldn’t be acquainted with it -, but in our descriptions. Actions in by-relation are events under special kind of descriptions, so they can have a causal description, under which the events no intentional character can have. Because by-relation is a relation with intentional character, it can’t be independent of our descriptions. There’s no events in the world, by-relating to others, but to reach some of our purposes, we must and can use these kind of concepts, as theoretical aids to explain some forms of intentionality.

Helena De Preester

Gent University, Belgium

Tamas Demeter

Trinity College, Cambridge / University of Miskolc

Dr Julien Deonna,

University of Cambridge, Department of History and Philosophy of Science

William Fish

Department of Philosophy University of Nottingham United Kingdom

A distinction has often been drawn between emotions – object-directed intentional states – and moods – emotion-like mental states which are not obviously object-directed (e.g. Searle in Intentionality). However, if such non-intentional moods exist, then this stands to refute the doctrine of intentionality – the claim that intentionality is the mark of the mental.

In this paper, I argue that there is no intrinsic difference between emotions and moods – at heart, they are examples of the same mental kind. However, I reject the claim that a correct theory of emotions / moods should proceed by assimilating either moods to emotions or emotions to moods. On the contrary, I argue that the traditional accounts of both moods and emotions are independently problematic. In the light of this, I suggest we require a new way of looking at this debate which takes into account the intuitive features of both emotions and moods. This paper begins that important task.

Taking my lead from Sartre (in Sketch for a Theory of the Emotions), and building on the intentional account of pain developed by Crane (in “Intentionality as the Mark of the Mental” in O’Hear 1998), I suggest that both emotions and moods are ways in which we are intentionally aware of objects in the world. However, unlike Crane’s analysis of pain (and in contrast to the more traditional account of emotions), I reject the claim that an emotion is a singular mental state linking the person with the purported intentional object of the emotion. This would seem to imply that an emotional engagement with an object is simply an “addition” – a state which can be added to or subtracted from an otherwise complete set of intentional states. However, I argue that being in an emotional states is far more “globally affective” than this account appears able to accommodate.

In its place, I develop an account of emotions and moods as “intentional modifiers” – states which are parasitic on our normal intentional engagement with the world and serve to modify the way in which we see the world. This approach enables me to retain the doctrine of intentionality as there is no scope for non-intentional consciousness – there has to first exist an object-directed intentional state in order for an emotion to subsequently modify it. Importantly, it is argued that emotional states serve not only to modify what it is like for us to intentionally experience the objects in the world, but that they also modify what we perceive to be rational, reasonable or normal behaviour to the objects in the world.

I then defend this account by showing how it deals with certain emotional occurrences which, it is claimed, are intuitively taken to be singular mental states linking the person with an intentional object. An example of this type of case is being angry with a particular individual. In such cases, it appears intuitive to hold that our anger is directed towards that individual, and is hence straightforwardly intentional. In response to this, I show how the theory of emotions as intentional modifiers can explain the apparent directedness of such states. This explanation not only serves to further clarify the theory itself, but also demonstrates why the traditional distinction between object-directed emotions and non-object-directed moods is not, in fact, required.

Anandi Hattiangadi

Trinity College, Cambrigde

There is a thesis about content that is so simple and intuitive that it seems undeniable. The thesis states that for a concept, F, to have any content, it must apply correctly to some things and not to others. Thus, for example, if F is the concept horse, it must apply correctly to all and only horses. If F applied correctly to all and only cows, it would be the concept cow rather than the concept horse. In general, the content of a concept can be given by the conditions of its correct application.

This apparently trivial assumption about content is widely reputed to imply that content is normative. At a first pass, this is to say that an agent’s grasp of a concept, her understanding of its content, guides her application of it. If she grasps the concept green, then she ought to apply it to all and only green things. If she fails to do so, perhaps because her vision is impaired, she has failed to do as she ought; she has made a mistake.

So understood, the normativity of content establishes a condition of adequacy for all theories of content. Any account of mental content must explain how it is that the appropriate patterns of correct and incorrect application get constituted. So, whatever account we give of content, it must tell us that an agent who understands the concept horse makes a mistake when she thinks that a horsey looking cow is a horse. This adequacy condition, it turns out, is not so easily met, particularly by naturalistic theories of meaning.

I argue in this paper that naturalists can avoid the problems posed by normativity, because the thesis that content is normative is untenable, at least as commonly conceived. It is not the case that I ought to apply the concept horse to all and only horses. First, I cannot be obligated to apply horse to all horses because ‘ought’ implies ‘can’ and there are more horses than I can think of. Second, I argue, we have no reason to think that I am obligated to apply the concept horse only to horses—though it may be that I can’t help but think lo, a horse in the presence of a horse. Moreover, by giving up the normativity of content, we do not thereby abandon the compelling thesis with which we began—that the content of a concept is given by its correctness conditions. The reason is that the compelling thesis does not imply the contentious thesis that content is normative. The normativity of content thus does not establish a condition of adequacy for theories of content.

Jussi Haukioja

Department of Philosophy University of Turku Finland

Melinda Hogan

University of British Columbia, Canada

A person can fail to understand a sentence. For example, she may not understand a sentence of Mandarin if she speaks only Russian. Also, a person can fail to understand a sentence in her own language; for example, she may not understand a sentence in her own language if the sentence has too many embedded clauses. And in other ways a person can fail to understand an utterance of a sentence in her own language. For example, even though she knows the sentence the speaker used, a hearer may not understand what is said if she has no idea why the speaker uttered that particular sentence. By contrast, in the case of mental representation, it may not seem so obvious that one can fail to understand a thought that one has. What would such a failure of understanding amount to, it might be asked? Thought tokens are not perceived, at least not in the way that public representations are. One does not see or hear them. One (normally) does not entertain hypotheses about what caused them, as one might entertain hypotheses about why a speaker chose the particular sentence he did choose on a given occasion. Some might continue in this vein (and have done so) to argue that we should not expect to find an answer to the question of what it is to understand one's own thoughts. We should not expect to find an answer to this question because the question is not well formed. Mental representations - understood as mental particulars occurring in one's own mind, and not as abstract propositions, are not the sorts of things that one can fail to understand. The view is expressed in the following line of reasoning: First, beliefs and desires and other mental representations are made up of concepts, which we may characterize as modes of presentation of objects. Second, there are good theoretical reasons for saying that concepts can be grasped in only one way, otherwise concepts will not do the work they are cut out to do. As modes of presentation, their job is to model the fact that there are different ways in which an object may be presented to one. Having such a model, we are able to explain an agent's behaviour in terms of the way in which she views the world. If there were different ways to grasp a concept in turn, we would not be able to explain an agent's behaviour in terms of her concepts. Evidence that would ordinarily count as evidence for the claim that two people view the world differently (e.g. evidence for the claim that they have different concepts of Venus, as the morning star and as the evening star) will no longer be evidence, if there are different ways to grasp a concept in turn. So, according to this view, there are not different ways to grasp a concept; nor can one fail to grasp a concept, for a concept (that is, a mode of presentation) just is a grasping.

I think this is a mistaken line of reasoning. It is possible to fail to understand a thought of one's own and to understand it in different ways over time. Therefore, an adequate account of mental representation needs to explain what is involved when one understands one's own thoughts. This is the main burden in the argument that follows. I will argue that despite the fact that there are good reasons to posit entities, mental particulars, designed to do the theoretical work of Fregean senses, there are equally good reasons from the point of view of psychological theory to make room for the fact that it is possible for a person to fail to grasp a thought she herself employs and to grasp the same thought differently over time. There is behaviour relevant to psychology that cannot otherwise be explained. I will also argue that despite the fact that linguistic representation - spoken language - provides no model for explaining understanding in the case of mental representation, in other quarters there is to be found a fairly straightforward way to explain what is involved in understanding one's own thoughts. I will make a suggestion about how to explain what is involved in understanding one's own thoughts. Finally, I will argue that the account of what is involved in understanding one's own thoughts that I describe has consequences for what it means to say that a thought is composed of concepts. In particular, it will not be essential to the identity of a thought that it be decomposable into a particular set of concepts. To explain what it is to understand one's own thoughts requires the idea of non-conceptual representational content.

Ferenc Huoranszki

CEU, Philosophy Department, Hungary

Informational semantics [IS] aspires to give a reductive analysis of intentional content by claiming that it is constituted by nomological relations between concepts and properties. IS has been offered as an analysis which represents an alternative both to the functional role semantics and to etiological theories of content. According to the first, intentional content is individuated by virtue of its functional-conceptual role. This, however, implies radical holism and relativism about content which is better to be avoided. According to the second, intentional content is determined with reference to its causal history. But this approach fails to do justice to the counterfactual supporting nature of intentional content and therefore cannot properly explain misrepresentation. In the talk while I grant the merits of informational semantics, I would like to challenge two of its main objectives. First, I will argue that the invocation of nomic connections does not explain much. And second, I wish to show that, properly understood, IS does not provides us with a reductive analysis of intentional content. Instead, it should be interpreted as a non-reductive, respond-dependent dispositional analysis of intentional content.

Juraj Hvorecky

Center for Theoretical Study, Czech Republic

A standard problem discussed in the 20 ^th century philosophy of mind is that of intentionality of non-veridical perception. Debates over the intentional nature of objects of hallucinations, imagery or visual illusions tend to take for granted that some role for the supposed objects of those perceptions is to be assigned and dealt with. Even in recent exchanges between phenomenalists (Block, Nagel) and representationalists (Tye, Dretske) this assumption goes unquestioned. In fact, from the respective viewpoints of both approaches, this is a fully justified step. Phenomenalists need to defend phenomenal consciousness as a special domain, distinct from intentionality of cognitive representations and/or neuronal processing. Entities appearing in phenomenal consciousness are naturally ascribed roles of phenomenal objects or events. Representationalists, on the other hand, need a unifying component, that would help them to propose an over-all theory of consciousness. According to them (see Tye 1995), representation is that element holding all forms of consciousness together and the main explanatory feature thereof. They therefore also see no problem in ascribing intentionality to all perception, whether it is veridical or not.

We believe that both of these strategies are running dangerously close to postulating full-blown phenomenal objects that sense datum theorists became infamous for. Alternative reconstructing of perception that only takes seriously veridical perception and its object is worthy exploring. Following Sellars (esp. Sellars, 1956/1997), we will discount intentionality of non-veridical objects as being fully dependent on normal vision. Appearance and impressions are according to this strategy explained in purely functional terms, as (neuro-) physical events serving the same functional role for the organism as usual vision without conveying anything about any objects. We will show that a plausible theory of consciousness – the main reason behind postulating intentional objects of non-veridical perception – can be worked out and plausibly defended. Unlike both of the above-mentioned theories, this one relies heavily on linguistic practice and behavioral evidence, standing closely to Dennett’s heterophenomenlogy (Dennett 1991).

Zoltán Jakab

NSERC postdoctoral fellow, Rutgers University

Certain representational externalists (F. Dretske, M. Tye, A. Byrne and D. Hilbert, W. Lycan) have recently launched an ambitious project of uniting the two fundamental philosophical problems about the mind: intentionality and phenomenal consciousness, into one: intentionality. These two problems have traditionally been viewed as separate, phenomenal character not being reducible to intentionality (nor the other way round). This tradition is now challenged by a view that we might call phenomenal externalism: phenomenal character, or the “what-it-is-like-to-undergo-them aspect” of certain mental states, is one and the same thing as a special type of intentionality. The subtype of intentionality in question is the aboutness of supposedly modular, non-conceptual, perceptual representations: their way of representing properties of external stimuli. On the most detailed versions of the theory, phenomenal character is either identical with, or supervenient on, the non-conceptual representational content of perceptual states. Representational content in turn is understood as something very close to information: lawlike causal covariation (between stimulus properties and the perceptual states that indicate them) plus, in some accounts (e.g., Dretske’s), as evolutionary function.

Phenomenal externalists have encountered stubborn problems in accounting for color experience, and most of these problems are empirical in nature. First, virtually all versions of the theory need the assumption that object colors are physical types of some sort, even though in fact this is not the case. Second, certain perceived attributes of the colors, like their similarity relations and unique-binary division, are not underlain by relevant stimulus properties the representation of which could explain how these perceived attributes arise. Third, individual differences in color perception make it the case that phenomenal color character varies independently of any stimulus information that color experience might carry, and this finding falsifies the identity or supervenience claims between (externalist) color content and phenomenal color character. Fourth, phenomena of simultaneous contrast make it extremely difficult to identify object colors with any inherent, local property of surfaces.

In my talk I propose to present my own version of some of these problems: the first, the third, and perhaps the second. I shall explain why these problems have fatal consequences for phenomenal externalism. I also think, on grounds outlined by Kathleen Akins, that these problems generalize to other perceptual modalities as well (Akins, in one of her papers, analyzes in detail heat sensation). For these reasons I shall suggest that the two traditional problems about the mind remain separate. The alternative general approach to phenomenal character is internalism: the idea that, for instance, what it is like to see colors is fundamentally determined by (i.e., supervenient on) how our visual systems operate.

In my opinion, sensation and perception can misrepresent the world in all kinds of different ways so long as the misrepresented aspects do no threaten survival. To achieve evolutionary success, not all features of the environment that affect perception have to be veridically perceived. To a certain extent, representational externalists are prepared to accept this possibility. However, it seems that there are too many systematically distorting aspects of perception (in addition to the veridical ones that I do not deny) to accept the idea that what it is like to undergo a perceptual experience is well explained by what observer-independent properties are represented (faithfully, as it were) by that experience. (Distorting, unfaithful representation is misrepresentation, and systematic misrepresentation in perception counts as “internally creating” one’s perceptual world – a process that is determined by how one’s mind operates, not in any way how the external world is.)

Greg R. Jesson

The University of Iowa, Department of Philosophy, USA

Frege’s ‘On Sense and Reference’, written in 1892, is widely considered to be a paradigm of Analytic methodology, yet curiously there is great disagreement on what it is about. This paper argues that Frege’s article is best understood as an account of intentionality that bridges the Continental and Analytic movements by applying new methods to old problems. Frege attempts to illuminate what Brentano called the “reference to a content” in every mental phenomenon by providing an explication of the cognitive difference between informative and non-informative identity statements. Although Frege does not use the word "intentionality", it is this pointing phenomenon of consciousness, as revealed in thinking and language, that drives Frege to offer his account of referring by means of Sinne to Bedeutungen . According to Frege every investigation into mental and linguistic acts must always be concerned with the sense/reference nexus--or more precisely the Thought/reference nexus.

Attempting to avoid psychologism, Frege insists that Sinne exist independently of the psychological lives of individuals and yet are involved in the explanation of how ideas and expressions, in mental and linguistic acts, have the referential character that they do. The difference between informative and non-informative identity statements cannot be accounted for by means of expressions and referents alone. Frege argues that the only way to explain the epistemic difference between these is by finding some way to bring the same object before the mind by means of different modes of presentation.

In this paper I claim that Frege's notion of Sinn , as lying between the psychological representations in individuals and the objects of reference, turns out to be deeply problematic. I argue that Frege faces a dilemma: Either Sinne are not as objective as he wishes, from which it follows that he lacks an account of why we are not marooned in the psychological realm; or Sinne are as fully objective as he maintains, from which it follows that they are not suitable to be the means by which ideas and expressions enjoy their referential capacity to the kinds of things we ordinarily take our mental and linguistic acts to be about. If Sinne are incorporated into the psychological realm to explain the intentionality of acts, then they will be incapable of accounting for objectivity; if Sinne account for objectivity, then there is no explanation of how our mental and linguistic acts refer to the kinds of commonplace objects we take them to be referring to. This dilemma is revealed by Frege in that Vorstellungen (subjective ideas) are required to accomplish many of the tasks of Sinne, and yet if Sinne are the instruments by which we apprehend Bedeutungen we seem removed from those objects in a way that is incompatible with our experience.

Concerning the first horn of the dilemma, Frege requires Vorstellungen to perform functions that are supposedly unique to, and are reasons for positing, Sinne. Vorstellungen have an explanatory advantage by being embedded in mental life but as Frege emphasizes, Vorstellungen could never account for the objectivity of knowledge because they are essentially private. However, the grasping of a referent is not possible without the mental or linguistic act grasping a Sinn . In 'The Thought' Frege claims, "Although the thought does not belong with the contents of the thinker's consciousness, there must be something in his consciousness that is aimed at the thought." At this level intentional non-existence is impossible—every idea grasps a Sinn. The objectivity of Thoughts precludes their isolation within consciousness, yet to avoid the skepticism Frege so rigorously attacks, consciousness must grasp something besides itself and its contents: images, ideas and impressions.

Concerning the second horn I argue that insofar as Frege holds that apprehension of Bedeutungen necessitates the grasping of Sinne, this results in puzzling philosophical consequences because it excludes direct access the very world Frege wanted to secure reference to in the first place. If reference to Bedeutungen is always mediated via Sinne, then it seems that we have only de dicto apprehension of the common objects of consciousness but de re access only to Sinne. Finally, I maintain that any theory that entails that we are incapable of directly thinking about common objects must be wrong. Frege's Sinne, like Hume's ideas, introduced to be the mediators whereby we gain access to the world, make it impossible to reach that world.

Stefania Ruzsits Jha

Researc Associate, PERC, Harvard University Cambridge USA

Michael Polanyi (1891-1976), a physical chemist of international repute and of Hungarian origin, is best known today for his writings on the standards and referee system in the scientific community, on the freedom of science, and especially for his concept of ‘tacit knowing’ in skillful performance.

Tacit knowledge is formulated as ‘from-to knowing,’ i.e. awareness directed from the subsidiary to the focal in insight, understanding and doing. His epistemology of Personal Knowledge was mistaken by logical positivists, Popperians and analytic philosophers to be a mystification of scientific discovery, rather than seen as an attempt to explain the non-explicit aspects of scientific insight.

Polanyi’s theory of tacit knowing shows influences of Gestalt Psychology, existentialism and phenomenology, culminating in an effort to work out a theory of meaning on a pattern developed from his conception of consciousness as vectorial, ‘from-to’ knowing. He called this vectorial nature ‘the functional aspect of tacit knowing,’ a relation between two kinds of awareness whose directedness is necessarily conscious. He borrowed the above notion of intentionality from Brentano, and made it the key feature of his theory of tacit knowing in his later works. Although he makes few references to the phenomenologists, he uses a few points from Merleau-Ponty to show how his theory goes beyond Merleau-Ponty’s notion of embodied knowing to working out the logical structure of tacit knowing. To build this structure on a scientifically objective foundation, he made use of relevant works of the phenomenologist physician Rothschild on biosemiotic theory.

Polanyi built his ontology on the foundation of his epistemology. The paper explores the phenomenological roots of Polanyi’s theory of tacit knowing. It also shows the connection between Polanyi’s conception of the act of knowing and his notion of ontological emergence, both of which are conceived of as vectorial. The author’s analytical approach makes Polanyi’s difficult language more accessible.

Tomis Kapitan

Northern Illinois University

Intention is the core concept in understanding human agency. In its terms we distinguish intentional behavior from other events, account for the causal influence of thought upon the world, explain the course of practical reasoning, and assess an agent’s responsibility. But to fully comprehend how intentions function in these explanatory roles requires articulating what the contents of intentional states are; what is it that we intend when we intend to do something? Naively, what we intend is to perform an action as distinct from a mere bodily movement. However, “action” is multiply ambiguous, and even if we speak of an action as a tokening of an action-type, we acknowledge that what is intended is not just any prospective event involving oneself. My conception of opening a door might include my thought of a door opening in response to my arm motion, but neither the door’s opening as a result of arm motion nor the motion itself is what I intend; these events could occur without my doing anything at all, for example, if my involuntary reflex causes a door to open. To capture what is involved, my content must include a representation of my own doing, my agency, my opening a door .

How is agency incorporated into intentional content? If causal origin is the distinguishing feature of agency, then it would seem that each intention embodies a representation of how a bit of behavior or its consequences is to be brought about. Since it is brought about by an intending, particularly, by the intending of the particular content in question, then the content of every intention would appear to includes a representation of its own causal influence and is, thereby, self-referential . When I intend to open a door, I aim to bring about a door opening because of that very intention to open it. In short, to intend to do an action A is to intend to do A as a result of that very intention. This way of embedding agency within intentional content is the self-referentiality thesis (SRT), advocated by philosophers such as John Searle and Gilbert Harman.

Some powerful arguments have been advanced favoring the SRT. One such argument assumes that while deliberating, an agent recognizes that intention-formation is a means to one's eventual action and, consequently, that in intending the action one thereby intends the recognized means.

However, this argument equivocates on the term “means.” In a broad sense, a means to an agent S's doing an action A is any necessary condition of S's performing A, yet the argument requires a more restricted reading where a means is any action that S must perform in order to perform A. No doubt one acquires an intention in order to act intentionally, but acquiring an intention is not always something one does; one can intend to do a certain action without having made any effort to put oneself into that state. Further, one can intend to undertake a course of action, say, to delete a file on a computer without deciding with which finger to press the button, even though one recognizes that one would not delete the file without moving a finger. Such basic acts are typically external to the action plan intended, and I submit that intendings—even if they are “doings”—fall into this category. Finally, even when we do factor an intending into an action plan, we do not thereby have a self-referential intending; it would be a prospective intending that is the component of the plan, not the present intending of that very plan itself.

The SRT also generates a vicious regress. When we impute causal efficacy to our own thoughts and intentions, we do so by way of knowing what it is we intend. Consequently, we must be able to independently identify an intention's content in order to determine its causal potential with respect to a given behavior. Because the SRT implies that one cannot intend to do an action A unless one intends that this very intending result in the action, then any effort to capture self-reference in the content of potentially efficacious intention is an endless task. Either it is impossible to finitely specify an intention's content, or the envisioned causal chain must terminate in a non self-referential intention. Consequently, either an intention's content is unthinkable or some intentions are not self-referential. Both alternatives spell doom for the SRT.

Caution must be exercised in determining the exact role of self-reflection within practical thought to avoid placing excessive demands upon the exercise of agency. We must distinguish the contemplative and active elements within reasoning about action, and thus, between our abilities to reflect upon actions theoretically, as actual or potential events within the world, and to consider them practically , viz., as that which is to be done. Only the latter are the proper objects of intention. When I intend to open a door, what I intend is just that, to open the door, and in so doing I give little heed to the causal mechanisms that underlie execution of this intention. Moreover, my content is not truth-valued. Instead, I think only of the agent (myself), the action-type, and a linkage expressed through the infinitive locutions “to”, “to do” or “to bring about”. The latter is distinct from the predicative linkage expressed by “am doing”, “does” or “will do”, and from the causal nexus expressed by '”brings about.” Through it, we can represent actions practically, and this is the closest we come to incorporating our own agency within intentional content.

Jean-Francois Lavigne 

Université Montpellier III (Pul Valéry)

"Husserl versus Searle, on Intentional and mental states.

Some aspects of a phenomenological criticism of J. Searle's conception of Intentionality."

In the introductory remarks of his "Essay in the philosophy of mind" entitled Intentionality (p.4), Sir J. Searle confesses that he willingly disregarded the contributions of the philosophical tradition on the subject, when elaborating his own theory of intentional states, as mental or brain processes. This paper intends to show that, at least concerning the major contribution of E. Husserl 's phenomenology to a careful and thorough analysis of the immanent structure of intentional states, this might well be a very unhappy and awkward choice.

The main purpose of this study is to put to a precise phenomenological test the basic and fundamental assertions of Searle concerning the "nature of intentional states" (Chapter I), by confronting them to the most original and acute functional distinctions in Husserl 's "description" of the general structure of intentional experiences ( "Akten", "intentionale Erlebnisse"). It is possible and much enlightening to show that, with as fundamental concepts as "real content" ("reeller Inhalt") in contrast to "intentional content" ("intentionaler Inhalt "), or the couple "Materie"/"Qualität " ("Matter" and "Quality"), the Logical Investigations of 1900-1901 already expounded several precious, decisive morphological and noetical distinctions which could have saved Searle's analyses and theses from many descriptive mistakes or shortcomings.

The examination takes its departure from Searle's fundamental thesis, that "Intentional states consist in representative contents under different psychological modes" (Ch. I, 3,3.). If one enters in a thorougher analysis of the structure of an "Intentional state" involved in a speech act, as functionally directed to its "intentional content", (more precisely to the "matter" involved in the "meaningful essence" (" Bedeutendes Wesen ") of this "act"), and without failing to see the functional and phenomenological difference between this "intentional content" and its "real content" ─ according to Husserl's clear demonstration in the fifth Logical Investigation ─ the first part of the quoted formula cannot but appear clumsy and regretably equivocal : Which "content" are we then speaking about ? Either it is the real content, and then it can indeed be viewed as "representative" or being a "representation", but as such it is no more "intentional" in the sense Searle means it ─ that is, in close connection with a meaningful, "speech" act ; or the expression means the "intentional content" as such, but then it is, more properly speaking, an intended content, the intentional or meaningful correlate of the whole real act ─ and then it cannot be characterized as "representative" any more, since it constitutes the meaningful core of the object represented.

After setting forth, in a descriptive-phenomenological manner, the necessity and objective value of those fundamental distinctions, it is also possible to argue in favor of them by putting under consideration the different morphological properties of both "contents"; here comes the ideal character of the "intentional content" to the fore, and plays a decisive role.

The general result of this first move is to show that, as soon as intentionality is at stake, and as far as it is concretely viewed, the traditional concept of "representation", and the oppositive scheme representation/object (or "state-of-affairs" naïvely conceived as an "object") are both unappropriate and utterly outdated .

My second main concern is with the second part of the chief thesis, meaning that the peculiar mode of subjective consideration or attitude under which a propositional content is expressed in a speech act (belief, doubt, wish, promise, interrogation, etc.) is a psychological mode : "A belief is a propositional content under a certain psychological mode". The lack of any distinction between the two modalities or forms of what can be equally called "belief" ─ between the psychical state of mind, and this objective-ideal mode of the expressed sentence, named by Husserl (after others!) "quality" ("Qualität") of the act ─ appears to be the main ground of Searle's psychologism. One only needs, then, to show that it is this originally psychologizing interpretation of the "Intentional state" that is supposed to fill the gap between intentional processes as subjectively experienced and the physicalistic reduction of mind to brain, to enhance the fundamental importance of this lack of precision, in the use of gross, phenomenologically and methodologically insufficient concepts.

A third point of examination will be the "simple", and perhaps slightly too naïve way in which J. Searle thinks it possible to cut straight in the long and far discussed question of the ontological status of intentional objects. Recalling the main features and contrasted positions of the rich controversy raised between Brentano , Marty , Twardowski, Meinong and Husserl on the problem, it appears no more possible to elude the difficulty ; one cannot simply disregard the numerous and tricky cases in which the "intentional object" is neither reducible to the "representative content", neither to a really existing object in the world, and shows nevertheless an objectivity of its own. I will show that deliberately ignoring this third form of objectiveness is what implies by Searle another shortcoming : merely presupposing the existence and structure of nature ("world") ─ the rightly famous "natural attitude" ─ that is, the most absurd contradiction one can find in epistemology and theory of knowledege.

Pierre Le Morvan

The College of New Jersey, USA

Intentionality is the property of being about, or directed at, something. Building on the work of Alston, Brentano, Burge, Dretske, Elgin, Kaplan, Quine, and Ryle, in this paper I distinguish three species of intentionality, which we may call transparent, translucent , and opaque respectively. I then show how drawing this threefold distinction makes an important difference in how a number of key theses and issues in the philosophy of mind in particular, and in epistemology and metaphysics more generally, ought to be assessed or even understood.

Philosophers in the analytic tradition often adduce propositional attitudes such as beliefs and desires as their only paradigm cases of intentional states. Hence, they tend to focus on one kind of intentional state (the opaque) to the neglect of two others (the transparent and translucent). By contrast, I begin my discussion with the intentionality of perception, of seeing in particular. Using distinctions drawn by Alston, Dretske, and Ryle, I differentiate objectual from factive seeing, and literal from figurative factive seeing.

I take objectual seeing (e.g., Sarah sees a rabbit) as a paradigm case of a transparent intentional state . A transparent intentional state meets each of the following conditions: (a) it is about, or directed at, something; (b) the ascription of such a state is existentially transparent (it licenses an existential generalization); and (c) the ascription of such a state is referentially transparent (it admits of substitution of co-referring terms salva veritate).

I take literal factive seeing ( e.g, Sarah sees that yonder object is a kumquat) as a paradigm case of a translucent intentional state. A translucent intentional state meets each of the following conditions: (a) it is about, or directed at, something; (b) the ascription of such a state is existentially transparent; and (c) the ascription of such a state is referentially opaque (it does not admit of substitution of co-referring terms salva veritate).

I take figurative factive seeing (where ‘seeing’ is used metaphorically, e.g. a blind boy’s seeing that (in the sense of realizing that or understanding that) Santa Claus is Kris Kringle) as a paradigm case of an opaque intentional state. An opaque intentional state meets each of the following three conditions: (a) it is about, or directed at, something; (b) the ascription of such a state is existentially opaque; and (c) the ascription of such a state is referentially opaque.

Having distinguished three kinds of intentionality in the case of seeing, I then extend the distinction to knowledge more generally. I show how one can capture, without the Russellian doctrine of sense-data, Russell’s distinction between knowledge by acquaintance and knowledge of truths. In this connection, I distinguish between transparently intentional objectual knowledge (e.g. Sarah’s knowing Martin), translucently intentional propositional knowledge (e.g., Sarah’s knowing that silver is the element with atomic number 27) and opaquely intentional propositional knowledge ( e.g., Sarah’s knowing that Santa Claus is Kris Kringle).

Capturing insights offered by Burge, Kaplan, Elgin, and others, I go on to apply the distinction between translucent and opaque intentionality to propositional attitudes such as beliefs, and show how there may be translucent and opaque beliefs.

The remainder of the paper is polemical. I argue that a conflation between the genus intentionality and one its species, opaque intentionality, lies at the heart of Wilfred Sellars’s case for the thesis that intentionality must be non-relational, Pendlebury’s case for the thesis that all perception is propositional, and Beckermann’s case for the thesis that biological tropistic systems cannot be intentional. I argue, furthermore, that once our threefold distinction is drawn, there are counter-examples to Brentano’s thesis that all and only what is intentional is mental, if we recognize biological tropistic systems as transparently intentional.

Finally, with regard to whether all conscious states are intentional, I argue that our threefold distinction sharpens the terms of the debate for both the proponents of an affirmative answer ( e.g., Lycan, Tye, and Crane) and for proponents of a negative answer (e.g. , McGinn, Searle, and Block). I argue that the following theses are straightforwardly false: (a) all mental states are transparently intentional, (b) all mental states are translucently intentional, and (c) all mental states are opaquely intentional. Hence, I argue that the genuine locus of the debate should be on the question whether all conscious states are transparently intentional or translucently intentional or opaquely intentional.

Jorge Louçã

DCTI / ISCTE Instituto Superior de Ciências do Trabalho e da Empresa Lisboa, Portugal

Brooklin College, CUNY, USA

When Brentano resuscitated the term ‘intentionality’ from Scholastic philosophy, he meant to use it as the ‘mark of the mental,’ a distinguishing characteristic by which to divide the world into mental and physical phenomena. Though analytic philosophy and cognitive science may dispute this contention, intentionality still plays a major role in their conception, analysis and explanation of mental states. It has become, in fact, a kind of holy grail of consciousness studies. Naturalists and physicalist's rightly view intentionality as the explanandum. If the ‘aboutness’ of mental states can be explained naturalistically, then the war is won. This is why Daniel Dennett, among others, addresses the subject at such length and in such detail. On the other hand Anti-reductionists, such as Searle, hold that intentionality is the one obstacle in the path of the reductionism which will not give way: intentionality is simply unamenable to naturalistic or physicalist explanations. Both camps often agree that intentionality is the right field of battle, since it avoids the nebulosity and imprecision of the other great explanandum of mental life: what David Chalmers calls the ‘hard problem’ of consciousness, or the fact that some brain/mental activities seem to be accompanied or constituted by awareness and self-awareness, while others are not.

Yet, as Husserl himself realized, “’consciousness of something’ is at one and the same time very obvious and highly obscure.” The intentionality of mental states—their ‘aboutness’ or ‘object-directedness’—proves to be a trickier notion than it at first may seem. A perception, thought, desire, or belief might inherently be ‘of’ something, but both the ontological status of the object intended as well as the relation between the mental act and the object are highly problematic. These issues are interrelated. I first examine the notion of intentional relations, contrasting them with physical relations. Here I show that while logical relations of “aboutness,” containment, and inference might bear superficial and metaphorical resemblance to physical relations of causation, spatial containment, and time-order, the difference between these mental relations and physical relations is wide enough to cast doubt upon the efficacy of a physicalist reduction of intentionality. The modern concept of physical causation, which limits the causal efficacy of an object to its shape, size, motion, and in more sophisticated versions, its role as a locus of forces, is particularly incapable of explaining intentional relations. Aristotelian or Scholastic causal theories of perception, I argue, are no better off. The acceptance of formal causes may at first seem to allow for a more amenable relation between thought, desire, perception, on the one hand, and the intentional object on the other. In perception, the form of an object enters my mind, leaving behind the matter. My perception is thus both caused by and “about” the mere form of a thing, which is something that already conforms to the notion of the mental. This, however, leads to obvious and insurmountable problem concerning the ontological status of the intentional object, in this case, the formal essences. I then show, however, that the modern causal theory of perception, though formulated in part precisely to avoid this problem, nonetheless fails to avoid it.

Lastly, I challenge the assumption that the problem of intentionality can be separated from the ‘hard problem’ of consciousness (as, for example, Dennett seeks to do). For this argument I go back to the original conception of intentionality found in the works of Brentano and Husserl, and, as I claim, Descartes. Descartes’ “objective reality” of an idea (i.e. its intentionality) is inseparable from the ‘formal reality’ of the idea. So, too, Husserl’s ‘noema’ is inseparable from his ‘noesis.’ The ‘aboutness’ of a thought, desire, belief, etc. is both conscious and self-conscious. For my thought to be ‘of’ something (even if non-existent), I must be aware both that is ‘of’ something and that it is my act of thought (with a particular phenomenology). Intentionality, I argue, implies both consciousness and self-consciousness. This is not jeopardized by the putative intentionality of propositions, sentences, computer programs, mechanical instruments, etc. In each case, I show that the attribution of intentionality is parasitic on a primary intentionality that cannot be separated from consciousness and self-consciousness. Intentionality and the ‘hard problem’ of consciousness cannot therefore be separated, and the former cannot be used to sidestep the latter. This is why attempts by Rosenthal to explain the internal intentional relation between intentional act and object intended by appeal to second-order thoughts (where the relation between first- and second-order thoughts is conceived of in causal terms) ultimately fail.

Brian McLaughlin

Rutgers University, New Brunswick,USA

I will propose a topic neutral analysis of color, and then make a case that color realism can be vindicated if we relativize colors to types of perceivers and types of circumstances.

I argue that if there really are colors, then colors are properties involving the potential interaction between matter and light. I offer a functional analysis of the concept of color, and then explain how it can guide vision scientists in their attempt to locate colors among physical properties. I explain how a color physicalist can account for the unity of colors – e.g., such facts as that red is more similiar to orange than it is to blue –, and for the unique/binary hue distinction. And, employing results from opponent processing theory – the leading neuro-computational theory of color vision – I say what I believe the most likely physical candidates are for being the colors that normal humans see.

Balázs M. Mezei

ELTE BTK, Budapest, Hungary

The problem of intentionality is minutiously developed in Husserl's various works. One strand of Husserl's conception can be found in his criticism of Locke's theory of knowledge. This ciriticsism is interesting for two reasons: first, through the Husserlian criticism we have a reliable and detailled summary of Husserl's conception of intentionality. Second, his conception of intentionality receives a particularly stimulating character through his ciritical reflections on Locke's central points concerning the making of human knowledge. In my presentation I shall not only summarize Husserl's relevant lines of thought, but I intend to point out too some of the weaknesses of his conception of intentionality. The most important defect of the Husserlian conception of intentionality can be found in what is usually labelled idealism: Husserlian idealism, it seems to me, cannot find an answer to the fact that human knowledge is organized along the lines of what we can call the bottom-top modell of knowledge as opposed to the top-bottom conception of epistemological idealism.

Ruth Garrett Millikan

Univeresity of Connecticut,USA

Teleological theories of mental content have generally been mistaken for theories of representation, but that is not really what they are Intentionality as "ofness" or "aboutness" is not explained by a teleological theory. Only falseness or failure to represent is explained. A teleological theory must always ride piggyback on some underlying theory of what constitutes real or true representation. I will discuss some relations between underlying theories that take basic representation to be like natural signing and theories that take basic representation to be like picturing, arguing that there is surprisingly little difference between these two if the nature of natural signs is understood correctly.

Bence Nanay

Department of Philosophy University of California, Berkeley

Jesse Prinz

Washington University in St. Louis

Some defenders of informational semantics present the theory as an alternative to causal history theories of intentionality. Informational semantics is also presented as an alternative to descriptive theories, resemblance theories, and phenomenal theories (those according to which consciousness is somehow essential for intentionality). It appears we have a variety of very different theories and must select between them. The latter three theories: descriptive, resemblance, and phenomenal theories enjoy comparatively little support in the present Anglophone philosophical climate. Causal history theories have gone out of vogue in the philosophy of mind, and they have been replaced by teleological theories, which appeal to Darwinian mechanisms. This talk attempts to reconcile some of these differences. If we adopt an informational semantics, I will argue, there is good reason to think that concepts are structured and perceptually based. This opens up a role for explanatory constructs that relate to (without quite being equivalent to) description, resemblance, and phenomenal qualities. I will also argue that informational semantics needs a causal history component. Causal history is needed to solve the disjunction problem that plagues informational theories. The only plausible version of the asymmetrical dependency approach to the disjunction problem covertly calls on causal history. Teleological theories, which sometimes purport to be special cases of the causal history view, are just special and unusual instances of it. If I am right, informational and causal history theories are complementary, and their marriage gives birth to the kinds of theories that both were designed to replace.

Howard M. Robinson

Central European University, Budapest, Hungary

The sense-datum theory of perception is generally taken as involving what is called a 'veil of perception'. This, in its turn, is thought to threaten scepticism, or, at best, to make our contact with the world counterintuitively remote. The thesis of this paper is that this is false. The 'direct realist' element in perception comes from the nature of its judgemental component, not from its phenomenal component; and the belief that sense-data cut us off from the world comes from the inadequate understanding of intentionality that the empiricists possessed, not from their doctrine of sensible content per se.

Mark Rowlands

Department of Philosophy University College Cork Cork, Ireland

Most accounts of representation are predicated upon three assumptions that collectively delineate the logical space within which debates about representation must, it is thought, be decided.

1. Representation is passive. A representation is the end point in a causal (nomic, teleological) chain originating with the object the representation is a representation of. Representation occurs when, as a result of such a process a representation is tokened in a subject. The representation can, of course, figure in further causal chains – ones, say, involving rational inference or motor action – but what makes an item a representation is its possession of a content that is determined by the representation’s place as the terminus of a causal (nomic, teleological) chain.

2. Representations are internal. Tokening a representation consists in the production of an internal configuration in a subject. This internal configuration is produced as the end product of a causal (nomological, teleological …) chain, and as such may be, individuation dependent on this chain or (more likely) some part of this chain. Nevertheless, external individuation does not entail external location. And the representation is located inside its subject, even if the content of this representation is externally determined.

3. Representation occurs at a time . Representation of X occurs at whatever time the representation of X is passively and internally tokened in a subject. Representation, that is, has genuine duration . While a representation can be possessed by a subject for an indefinite period of time, representation of this fact (in that subject) occurs only when the representation is activated. And the activation of a representation is something that has genuine duration. It begins and ends at definite (although perhaps difficult to determine) times, and has no intermittent lacunas. Representation is, thus, essentially synchronic, rather than diachronic.

This framework defines what is to count as naturalizing representation. To naturalize representation is to explain the relation that connects internal representation with external item represented. This explanation must, at the very least, explain why the representation is about the item it represents. However, must do this in ways that presuppose no representational concepts, or progressively in way that presupposes only less than fully representational concepts. Accordingly, what often passes as an explanation of representation is the provision of a non-intentional criterion for a representation being connected to one item rather than another

Against the traditional framework for understanding (hence naturalizing) representation, I shall argue for three counter-theses:

1a. Representation is active. Representation can consist not in the production of a passive representation but in the active probing and exploration of environmental structures.

2a. Representation is (partly) external . The probing and exploration of environmental structures is an activity that takes place not inside the subject’s head but in the subject’s world. While such probing is clearly supported by operations occurring inside the head of the representing subject, the exploratory activity is itself an external or environmental activity.

3a. Representation occurs through time . Representation of a given environmental contingency is not the sort of thing that can occur at a time. Such representation consists in, (at least in part) the probing and exploration of environmental structures, and such exploratory activity is something that takes place through time rather than at a time.

Focus of discussion will be provided by the Milner-Goodale dual visual system hypothesis, and work on change blindness associated with Kevin O’Regan.

Dan Ryder

Department of Philosophy, University of North Carolina

There are many different kinds of representations: pictures, maps, words, graphs, diagrams and more. I present evidence that the kind of representation that typifies the cognitive/perceptual part of the brain, the cerebral cortex, is model representation. A model M is a structure that represents another structure E because M is designed to be isomorphic to E. M has the function of being isomorphic to E, or mirroring E, and the structural nodes {m₁, m₂, m ₃ …m _n } of M have the function of corresponding to the structural nodes {e ₁, e₂, e₃…e_n } of E. When this is the case, we say that {m₁, m₂ , m ₃ …m _n} stand in for and so represent {e ₁, e ₂ , e ₃…e_n}.

Models can be produced by machines. A simple example would be a machine that takes an object as input, produces a mould from that object, then shrinks the mould and eventually produces a miniature model of the original object from that mould. In this case, there is no intentional agent who designed the model directly, so in deciding what the miniature represents, and what its parts have the function of corresponding to, we cannot consult the intentions of such a designer. Instead, we consult two things: 1) the general operating principles according to which the model building machine was designed, and 2) the object that acted as a template for the machine to produce the model in accordance with its operating principles. Thus if both the original object and its miniature model come to have a gouge in corresponding positions on their left sides, but only by accident, we do not say that the gouge on the model represents the gouge on the original object. Only features of the model that are produced by the model building machine from the original object in accordance with its general operating principles represent the corresponding features in the original object. This is not to say that a model building machine can produce only perfect models. On the contrary, the machine’s operating principles may face limitations in certain situations, such that although the machine is functioning according to its operating principles, it may produce an imperfect model.

Recent evidence suggests that the cerebral cortex may well be a very general purpose model building machine, designed through natural selection to produce dynamical models under guidance from the environment. I outline the principles under which the cortical model building machine may operate, based upon the SINBAD theory of the cortex. These principles allow us to pick out what kinds and individuals served as “templates” for the cortex to produce a model of some portion of the environment, and thus to determine what particular cells or groups of cells in the cortex have the function of corresponding to, i.e. what they represent. (Unlike in the simple mould-plus-shrinking device described above, it is usually a determinate matter whether it is an individual or rather a kind that has guided the cortex to produce a particular model.) Since this mechanism would develop representations of just the sorts of things that we represent, there is reason to suppose that it reveals the nature of mental representation, at least in our kind of mind.

A dynamical model of the sort that the cortex constructs is a useful thing. Since it operates isomorphically to some dynamical structure in the environment, it can “fill in” missing information when that information is not directly available. For example if the thermometer in the Space Shuttle’s hull is broken during atmospheric re-entry, a model of the Space Shuttle may be used to discover what the hull temperature is simply by plugging the known variables (velocity, atmospheric density, angle of descent etc.) into the model and reading off the temperature from its corresponding node in the model. This “filling-in” ability of models may also be used to discover the best action to take in a particular circumstance. For instance, one could discover the velocity below which one would need to travel in order to avoid exceeding a certain maximum hull temperature. One would simply plug in that hull temperature, and the atmospheric variables, and read off the maximum velocity from its corresponding node in the model. These uses of a model provide the basis for an account of both occurrent belief (i.e. judgement) and occurrent desire.

Susan Schneider

Rutgers University

There has recently been much interest in the view that psychological explanation should be wide, that is, the view that psychological kinds fail to supervene on the intrinsic states of the individual. Theories of wide content claim that content is individuated externally; theories of broad content, (as I am using the expression), in addition to this, take the basic semantic properties of thoughts to be denotation and truth. Following Jerry Fodor, let us call the view that broad content is the only sort content that individuates psychological kinds, “Broad Psychology.” Broad Psychology can be adopted by those sympathetic to Russellianism, the semantic view which claims, inter alia, that attitude ascriptions that differ only in coreferential expressions have the same meaning. The Russellian about psychological explanation accepts this semantic view and adopts, in addition to this, the following claim about psychological explanation: beliefs differing only in containing coreferring expressions, although they may very well differ in mode of presentation, are to be treated by intentional psychology as being type-identical and are thereby subsumable under all the same intentional laws.

Frege's puzzle about belief ascription is a well-known problem arising for the purely semantic version of Russellianism. Along similar lines, the Russellian about psychological explanation faces a related worry, the problem of what I will call, following Fodor, "Frege cases." Frege cases involve agents who lack knowledge of certain identities, where such knowledge is relevant to the success of their behavior, leading to cases in which the agents fail to behave as the broad intentional laws predict. To cite a well-known example, Oedipus threatens to be a counterexample to the broad generalization: “(M) Ceteris paribus, if people believe that they shouldn’t marry Mother and they desire not to marry Mother, they will try to avoid marrying Mother” because, in virtue of trying to marry Jocasta, according to a broad psychology it is also true that he tries to marry Mother. In this way Frege cases present Broad Psychology with putative counterexamples to its laws.

It is generally agreed that Frege cases are a major problem, if not the major problem, that this theory faces. And the general view is that Broad Psychology is not faring well with respect to this problem. However, I believe that this line of thinking is mistaken, and in this talk, I will proceed to outline a new solution to the Frege cases.

Many have criticized Fodor’s recent attempt to justify including Frege cases in the ceteris paribus clauses. To justify including them in the clauses, Fodor appeals to an epistemological argument in The Elm and the Expert. [Elm/44-50] But it strikes me that when critics go to such great lengths to attack this justification they are failing to speak against a second possible justification that arises when one considers some larger issues. In fact, this justification is more parsimonious than Fodor’s, in the sense that it merely invokes argumentation that the broad psychologist needs to make in any case.

I will argue that including the cases in the ceteris paribus clauses is justified by a larger theoretical decision for intentional laws with a canonical form that is broad. Any decision to include them in the clauses is the result of an overall assessment of the debate about which canonical form intentional laws should take, broad or narrow. While such a decision is a global affair, I shall focus on the part of this theoretical decision that I believe is internal to Frege cases – the part which involves the issue of whether computational explanation of coreferring thoughts will suffice.

The literature has been quite negative on this score. Against these critics, I will argue the following: first, Broad Psychology can treat Frege cases as tolerable exceptions, rather than counterexamples, to broad intentional laws. Second, there is no missed prediction of Frege cases: Broad Psychology does not fail to explain events in its laws that narrow psychology, on the other hand, captures (under narrow description). Further, although Broad Psychology must include Frege cases in the clauses it can still offer certain generalizations to predict Frege cases. And finally, I will respond to claims that explanation of Frege cases must be intentional that are based upon the general concern that only narrow intentional explanation can explain what happened to Oedipus, and alternately, upon a skepticism about interpersonal typing of modes of presentation.

Erwin Tegtmeier

University of Mannheim, Germany

János Tõzsér

Institute for Philosophical Research Hungarian Academy of Sciences

The Content of the Perceptual Experience

(Arguments for the Disjunctive Theory)

In my paper I would like to defend the following claim:

In arguing for the above thesis I wish to accept the disjunctive theory of perception. This means that if it is true that we perceive the mind-independent world directly, we shall have to reject all theories according to which the same mental state is shared by a veridical perception and the corresponding hallucination.

I am definitely in agreement with John McDowell, according to whom in the case of a veridical perception the object of perception is a constituent of the perceptual experience. Therefore, it is clear that we cannot be in the same mental state in veridical perception and in a phenomenologically indistinguishable hallucination.

I think that in order to be able to reason successfully in favour of the disjunctive theory of perception, I should demonstrate that other theories of perception – in spite of appearances – are not able to explain the claim that we have direct access to the world.

It is clear that the sense data theory is incapable of achieving this, but it is far from trivial that neither is the intentional theory, however sophisticated it may be. In a large part of my paper I will argue that the intentional theory is also unable to explain the transparency of perceptual experience.

Whoever accepts a common factor view (including the intentional theory of perception), must admit that someone could be in a given state of mind, however, the world should not be in the way that it is presented as being. In a word, the state of mind is not by itself sufficient to guarantee that the world is in a certain way. If intentional theorists are right, then our cognitive powers are exhausted by having such experiences. I shall argue that this view has the consequence that our cognitive powers do not reach all the way to the world, and there is a gap between the mind and the world.

Shannon Vallor

University of San Francisco, USA

The Intentionality of Reference in Husserl and the Analytic Tradition

The last century has given rise to many innovative and compelling approaches to doing philosophy. Viewed separately, the analytic and the phenomenological traditions represent perhaps the most promising and notable examples of these remarkable new directions in thought. Viewing them side-by-side, however, we are confronted with the rather embarrassing fact of how little, historically, these traditions have had to say to one another. In recent decades, the efforts of a small but growing contingent of committed thinkers to lead these traditions into genuine dialogue have begun to pay real dividends. Yet much still remains to be done.

In this paper I argue for the importance of this exchange by demonstrating some of its concrete benefits. Using the phenomenon of ‘reference’ and some of its associated problems and paradoxes as an example, I argue that the analytic tradition in particular needs to participate in this exchange if it hopes to remain philosophically viable. One important reason for this lies in the divergence of the analytic and phenomenological interpretations of intentionality, which can ultimately be traced back to the early exchanges between Frege and Husserl. I argue that Frege’s unwillingness to recognize the distinction between the psychological and the intentional led him to construct a theory of reference which was doomed to failure. There is a long list of problems associated with Frege’s theory of reference and its analytic heirs, including Russell’s Paradox, referential opacity, difficulties with substitution, naming and identification, and the puzzle of informative identity statements. These problems, particularly Russell’s Paradox, eventually led Frege to conclude that his logical system was fatally flawed. Russell, Quine and others carried on, but in the direction of an artificial and barren formalism which never satisfactorily explained the phenomenon of reference, and ultimately dismissed it as a mere ‘mentalistic excrescence’.

Husserl, on the other hand, by acknowledging the nature of reference as an intentional act, was able to uncover an essential feature of reference that Frege overlooked. This is the directionality which characterizes all referential acts. In this paper I suggest how this feature can be used to explain much of what Frege’s heirs could not - for instance, why failures of substitution often occur in modal, demonstrative and reported contexts. One interesting consequence of this is the suggestion that Leibniz’ law of substitutivity, accepted by most analytic thinkers at face value, does not in fact apply to ordinary reference. The following example illustrates one case in which the analytic tradition’s unwillingness to acknowledge the act-character of intentionality weakens its grasp of the essential structures of phenomena it attempts to analyze.

The directional component of referential acts, combined with Husserl’s analysis of the noematic structure of intentionality, can give us a new understanding of how statements like ‘The evening star is seen in the early dawn’, ‘The loser of Waterloo was victorious’, and ‘Barbarelli was so called because of his size’ can result from seemingly legitimate substitutions. We need not resort to Russell’s clever but problematic theory of descriptions to solve these problems. We only need to recognize that in each case, the directional ‘track’ or ‘path’ of the original reference has been altered by the substitution, despite the fact that this path terminates in the same object as before. For example: The statements “Napoleon was victorious” and “The loser of Waterloo was victorious” refer to the same entity. They also both predicate the same thing about that entity. So why doesn’t the second sentence make sense? Why should it matter to the truth of the statement which name we use? Russell would have us believe that in fact we are not actually referring or using true names in either case (though we believe that we are), but rather making existential statements containing disguised descriptions.

The answer is actually much simpler than that, and much more consistent with common sense. The two statements differ only in the particular path of sense they take to their referent. The key to this is that if references do bear within them a particular direction to their destination, then names are like route maps to that destination. Just as there can be many possible routes to a destination, an entity can have many names. In the first statement, the name ‘Napoleon’ indicates an intentional interest in a particular group of determinations or properties belonging to the referent. ‘The loser of Waterloo’, though not a proper name, functions the same way in the second statement. But in this case it indicates a different direction of interest and a correspondingly different ‘path of sense’ to the same referent. The statements share a common terminus - the entity, or in Husserl’s terms, the identical ‘noematic X’, which both sets of sense determinations have at their core. The problem occurs when a statement is altered by substitution to refer by way of a new path, which now travels through a sense determination (‘loser’) which is contrary to that predicated of the object (‘victorious’). Yet this only happens in artificial instances, when a philosopher treats the statement not as representative of a living intentional act, but as a passive construct. Ordinary references don’t encounter such problems because they are acts, intelligently formed. Unrestricted substitution in ordinary language works as well as mapping two different routes to your destination, then swapping the names of the landmarks on one map for those on the other. Only someone who didn’t understand how maps work would think that because the maps direct you to the same place, the points along the way ought to be interchangeable. Thus the directional component of intentionality is reflected in language, so that when we speak we are not merely labeling and describing objects in the world, but expressing to others the specific nature and focus of our interest in them. Any analysis of language that hopes to avoid aporia and paradox must address this component.

Rob Vanderbeeken

Centre for Logic and Philosophy of Science Ghent University Belgium

This paper argues that intentional behavior should be explained by means of Causal Mechanical explanations that refer to the mechanisms of practical reasoning, stating the causally significant beliefs, desires, intentions, the used rules of reasoning and the like. I believe that such CM explanations are efficient and unproblematic, unlike action explanations using Ceteris Paribus laws or counterfactuals.

A general distinction between three methods of explanation can be made: nomic, modal and ontic explanations. Nomic explanations are arguments (cf. Hempel and Oppenheim [1948] (1965)). Davidson (1980) argues DN explanations of actions face the problem that there are no strict psychological laws. Nevertheless, Fodor (1991) and Pietroski (2000) defend the idea that cp-laws can do the job. Modal explanations use conditionals stating necessary and sufficient conditions (cf. Lewis (1973)). Due to the flaws of the DN approach, several philosophers consider modal explanations to be the only option for explaining actions (e.g. Von Wright (1971), Rudder-Baker (1995)). Ontic explanations, on the other hand, are singular explanations that refer to the causal mechanisms that provoked the explanandum. Salmon (1998: 324) distinguishes two sorts of CM explanations: etiological explanations tell the causal story leading up to the occurrence of a phenomenon, while constitutive explanations provide a causal analysis of the phenomenon itself, referring to the underlying causal mechanisms that constitute the phenomenon. When it comes to actions, most philosophers only take the latter sort into account: CM explanations are always considered to be reductive explanations that state underlying brain states or neurophysiological mechanisms. I believe this unilateral view is a missed opportunity: etiological CM explanations, using psychological concepts, are the best candidate for action explanations.

Section 1 argues that nomic explanations are inappropriate instruments: even if cp-laws were unproblematic, we don’t need laws for explaining actions. Section 2 discusses modal explanations by means of counterfactuals. Some philosophers recognize that counterfactual theories of causation are problematic, but still they believe that counterfactual theories of causal explanation escape these problems. Unsurprisingly, this a mistake. My overall argument against counterfactual explanations is that, even if they were unproblematic, they are unnecessary complex for explaining actions. In both cases of nomic and modal explanations of actions, one can only overcome the mentioned shortfalls by introducing causal assumptions. The remedy itself is sufficient for explaining actions, however. I prefer to skip the detour using cp-laws or counterfactuals and focus unequivocally on the relevant causal claims. Section 3 discusses my account of etiological CM explanations of actions. Two central claims are the following: (i) CM explanations of actions have different formats. If the explanation seeking why-question is a request for the explanation of a contrast, stating a particular belief, for instance, can be an adequate explanation. If a why-question requests the explanation of a fact, an adequate answer requires a full description of the practical reasoning involved. (ii) The used rules of reasoning (e.g. a rule of thumb, an accepted norm, a decision rule) can be considered as a kind of propositional attitudes that, if made explicit, can be referred to as an explanans. By providing the opportunity to take all sorts of rules into account, we obtain a model for explanation that allows for a wide range applications, unlike Hume’s classical Belief-Desire model.

Agustín Vicente

Universidad de Valladolid (Spain)

Some philosophers (e.g. Antony and Levine (1997), Sturgeon (1998)) hold that the basic reason to naturalize intentionality is that the instantiation of intentional properties bring about changes in the physical/natural world, even though the physical/natural world is causally closed. Perhaps it is not true that this is the “big” argument for naturalism, but still it is true that the instantiation of mental properties seem to bring about changes in the natural world, and so it is true that a good naturalistic explanation of intentionality should explain this fact.

On the other hand, the defining feature of intentional properties is their representational character, or their being about. Intentional properties represent the world as being in a certain way, and, what is more, thay may err, if the world is not the way they represent it to be. That is, intentional properties may represent correctly, but also they many times misrepresent. This is a very striking feature about intentional properties, and it is, for sure, the one that shows, or has shown, more resistance to being naturalized. However, it is a fact that any naturalistic program must explain.

My purpose for this paper is show that these two characteristics of intentional properties (their being causally efficacious and their being susceptible of misrepresenting) have not been explained – both of them “at the same time” - by any naturalistic account. On the one hand, most philosophers who are focused on mental causation do not even take into account the problem of error, and treat intentional properties just as any “high level” properties (be they geological, biological, or whatever), that is, properties that have no representational content. On the other, we find that the most promising explanations of the problem of error construe intentional properties as teleological. However, teleological properties are, as such, causally epiphenomenal (as Millikan (1993) recognizes): a can-opener does not open cans in virtue of having been designed to open cans, but in virtue of its shape and its having sharp boundaries.

Only Fred Dretske, to my knowledge, has tried to give an account of intentionality that covers these two facts. His case dramatically exemplifies the tension existing between explaining mental causation and explaining mental representation. I do not have the space to develop Dretske’s account here, but, only to get the flavor of the theory and the problem: Dretske says that neural structures in the brain of organisms indicate (are activated by) properties in the environment. Now, the detection of some of these properties may be of special importance for the organisms (e.g. the organism is a rabit and the property eagle ), and so the neural structures indicating them get connected to the motor system, so that every time a structure is activated a motor reponse takes place (the rabit runs away). Dretske says that when the neural structures are connected to the motor system, they are recruited to indicate the presence of the property they indicate. That is, they acquire the function of indicating those properties. At the same time, it can be said that every time the motor response the activation of the neural structures generate takes place, it takes place because those neural structures indicated those properties, i.e because of the information they carried. Now, Dretske presents this as a theory of mental causation, but, as many critics have said, it cannot be so, for what he shows is, at most, that indicational properties are causally efficacious. Yet, indicational properties are not representational properties: a representational property, on his account, is a property that has been recruited to indicate another.

The first consequence of the existence of this apparent tension between causation and representation is that no actual or predictible naturalist program is complete. A second consequence, somewhat lateral, is that the argument from mental causation cannot be “the” argument for naturalism: otherwise one would never be disposed to drop mental causation from her naturalist account (as Millikan does).

Now, can I also conclude that there is no way to explain causation and representation (and so to explain intentionality) in a naturalistic way? I do not have any conceptual argument that would support this conclusion. However, I take it that it is very difficult that we come up with a theory of representation that improves over contemporary teleological theories. So the pessimist position is here the default position.

Alberto Voltolini

University of Eastern Piedmont at Vercelli

1. Philosophers of the Oxford variety used to solve metaphysical problems by means of linguistic analysis. Yet metaphysical questions should be answered by metaphysics and not by language. In this respect, the problem of the criteria of intentionality is an interesting test. Many philosophers have presented a linguistic analysis of such a problem. Yet this analysis entails eliminativism with regard to intentional objects. However, the linguistic criteria of intentionality should share the same prima facie ontological commitment as the non-linguistic criteria. More precisely, linguistic criteria should preserve two ontological intuitions that are expressed by the non-linguistic criteria, namely that ‘there are’ intentional objects towards which mental states are directed and that intentional objects are such, that they present themselves as perspectival entities. If such ontological intuitions are correct, they should be an object of enquiry for a metaphysician, not for a language analyser.

2. Traditionally, the first linguistic criterion of intentionality is the failure of existential generalization in attitude contexts. This failure entails that the existential quantifier occurs there in a narrow scope position. Yet we claim that the genuine linguistic counterpart of the first non-linguistic criterion is a wide scope occurrence of the existential quantifier with no existential implications. Hence, the singular term figuring in an embedded position in the non-quantified version of an attitude context has a wide scope occurrence as well.

3. Traditionally, the second linguistic criterion of intentionality is referential opacity. Yet the above claim entails that the genuine linguistic counterpart of the second criterion is what we label pseudo-opacity. According to it, ordinarily co-referential singular terms are unreplaceable salva veritate in attitude contexts even though they have a wide scope occurrence. In fact, in such a case they have a distinct reference which is not their ordinary co-reference. Let us call this thesis Frege’s thesis. We indeed distinguish between ordinary and extraordinary reference of a singular term, meaning by the latter the reference to an intentional object. Obviously Frege’s thesis entails that two extraordinarily co-referential singular terms are substitutable salva veritate.

4. According to a widespread opinion, referential opacity is not a sufficient criterion of intentionality, for it also characterizes modal as well as deontic contexts. We also claim that it is not even a necessary criterion. On our view, opacity strictly concerns ordinarily co-referential terms only when they occur in narrow scope. In fact, in such occurrences those terms not only lack of an ordinary reference, but they also fail to have an extraordinary one. The truth-conditions of a sentential tokening containing one such occurrence require that its embedded sentential part is true in possible worlds different from the actual one. This entails that the above occurrence has a possible denotation. Yet such a denotation is ontologically indeterminate. For it is not possible to assess whether the denotation of such an occurrence in a certain possible world is the same as its denotation in another possible world.

5. As we have seen, failure of existential generalization in attitude contexts means a narrow scope occurrence of the existential quantifier as well as of the relevant singular term figuring in the non-quantified version of an attitude context, and a narrow scope occurrence of such a term means referential indeterminacy. As a result, if these were the linguistic criteria of intentionality, intentional objects would be eliminated from the very beginning. On the contrary, our version of such criteria allows to prima facie preserve them as entities one may both refer to and quantify over. However, this says nothing about what either these objects or intentionality itself really are. Indeed, it is not only the case that a lot of ontological doctrines on intentional objects are compatible with our reconstruction of the linguistic criteria. Moreover, also different ontological stances on intentionality itself are possible, interpretationist as well as eliminativistic ones included.

Kenneth Williford

Department of Philosophy, The University of Iowa ,USA

The chief contribution of the Phenomenological tradition was the thesis of intentionality according to which all consciousness (or, at least, consciousness in its most characteristic form) is consciousness of something. Though many in contemporary analytical philosophy of mind accept some version of the thesis, there are others who are quite willing to deny it. In this paper I critically examine a representative externalist theory of content according to which the thesis of intentionality need not be true.

If an externalist theory of content implies, as some arguably do, that we do not know or are often wrong about the content of our intentional “states” as that content is given to consciousness , the proponent of the theory has a number of options open to him. Apart from abandoning the theory, the theorist can hold that consciousness, though intentional, is an unreliable guide to its real content and that it is left to theory to determine what that real content is. Or, more radically, the theorist can deny that consciousness is intentional.

If the externalist maintains that consciousness is an unreliable guide to its real content, she must 1) explain the epistemic unreliability of consciousness in this regard and 2) offer a theory, presumably in externalist terms, of the mistaken content that consciousness offers up. There are, however, considerations of a naturalistic sort that strongly suggest that an epistemically unreliable consciousness even when it forms reflective judgments about its own intentional content would be largely maladaptive. Moreover, it is doubtful that the types of externalist theories that raise this problem have the theoretical resources with which to offer a theory of the (ex hypothesi) mistaken content consciousness offers to itself. If the externalist theorist cannot provide such a theory, then she is forced to admit that her account of intentionality is, at best, incomplete.

The externalist theorist can escape from the above difficulties by denying that consciousness is intentional. The obvious difficulty for this option is that it seems to be a denial of a very basic phenomenological datum: consciousness seems to be intentional. The externalist theorist who takes this option will either have to deny that anything is given to consciousness or will have to admit that consciousness seems to be intentional but only so seems.

If the theorist takes the first route, then he must still offer a theory that explains why so many earnest people (philosophers and others) make judgments that, prima facie, seem to have “would be” phenomenological data as their content. That is to say, once again, the externalist must offer a theory of systematically erroneous content of this type. Moreover, the offered theory must be more plausible than the “theory” according to which phenomenological data are real. If the externalist takes the second route, then he will have to explain how it is that consciousness seems to be intentional even thought that very seeming is not itself intentional . This seems even less promising to put in mildly.

But there is deeper problem facing the externalist whose theory implies a rift between content as determined by theory and content as determined by consciousness. As is the case with any epistemically respectable theory, the theory of content must begin with empirical data. In the case at hand, it just so happens, some of the important initial empirical data is phenomenological. If one’s theory of content implies that in general the deliverances of one’s own consciousness are not reliable indicators of content, then one not only seems to dispense at once with one’s initial theoretical motivation and evidence base, one is also threatened with the self-undermining possibility that one does not even know or may be wrong about the content of one’s own theory of content. Less paradoxically stated, that we have generally reliable (though not infallible) conscious knowledge of content seems to be a presupposition of rational discourse. Therefore, any acceptable theory of content must explain why this is so. If a theory does not have the resources with which to do this, it must be abandoned. Based on the above considerations, I argue that many externalist theories do not have the necessary resources.

I conclude by outlining a theory of content that has an important externalist component, that involves no “non-natural” ontological commitments, that is faithful to the phenomenological data concerning the intentionality of consciousness, and that has the resources with which to explain how it is that our consciousness of our own intentionality is generally epistemically reliable.