Antirepresentacionalismo jmroy

36
ANTI-CARTESIANISM AND ANTI-BRENTANISM: THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM Jean-Michel Roy ABSTRACT: Despite its internal divisions and the uncertainty surrounding many of its foundations, there is a growing consensus that the on-going search for an alternative model of the mind finds a minimal theoretical identity in the pursuit of an anti- Cartesian conception of mental phenomena. Nevertheless, this anti-Cartesianism remains more or less explicitly committed to the neo-Brentanian idea that intention- ality is an essential feature of the mental—an idea that has prevailed since the advent of modern cognitive science in the 1950s. An issue of compatibility is thereby raised, as neo-Brentanism arguably sides with cognitive Cartesianism. The main goal of the paper is to put into full light one specific aspect of this largely unper- ceived problem of compatibility by arguing that the neo-Brentanian property of intentionality is an essentially representational one that runs counter to the salient anti-representationalism of anti-Cartesianism. And, that this representational essence confronts the search for alternative models of the mind of an anti-Cartesian kind with the following theoretical issue: To what extent is it possible to devise a non- Brentanian property of intentionality, particularly one that is fully dissociated from the property of representation? This issue is shown to be much deeper and more dif- ficult than it looks once the nature of representation is properly apprehended; it seems to be still waiting for an answer in the current search for an alternative model of the mind, if only because it has not yet be set in fully adequate terms. 1. ANTI-CARTESIANISM AND NEO-BRENTANISM: THE PROBLEM OF COMPATIBILITY The search for alternative models of the mind has been going on for some time now, but it remains arguably inconclusive. The assessment of what it has achieved so far very much depends, however, on the way it is to be Jean-Michel Roy has a main affiliation with Ecole Normale Sup erieure of Lyon (Lyon) as Professor of Philosophy, France, and a secondary affiliation with East China Normal Uni- versity where he is a recurrent visiting Professor in the Philosophy Department. The Southern Journal of Philosophy, Volume 53, Spindel Supplement (2015), 90–125. ISSN 0038-4283, online ISSN 2041-6962. DOI: 10.1111/sjp.12125 90 The Southern Journal of Philosophy Volume 53, Spindel Supplement 2015

description

Antirepresentacionalismo jmroy

Transcript of Antirepresentacionalismo jmroy

ANTI-CARTESIANISM AND ANTI-BRENTANISM: THEPROBLEM OF ANTI-REPRESENTATIONALIST

INTENTIONALISM

Jean-Michel Roy

ABSTRACT: Despite its internal divisions and the uncertainty surrounding many of itsfoundations, there is a growing consensus that the on-going search for an alternativemodel of the mind finds a minimal theoretical identity in the pursuit of an anti-Cartesian conception of mental phenomena. Nevertheless, this anti-Cartesianismremains more or less explicitly committed to the neo-Brentanian idea that intention-ality is an essential feature of the mental—an idea that has prevailed since theadvent of modern cognitive science in the 1950s. An issue of compatibility is therebyraised, as neo-Brentanism arguably sides with cognitive Cartesianism. The maingoal of the paper is to put into full light one specific aspect of this largely unper-ceived problem of compatibility by arguing that the neo-Brentanian property ofintentionality is an essentially representational one that runs counter to the salientanti-representationalism of anti-Cartesianism. And, that this representational essenceconfronts the search for alternative models of the mind of an anti-Cartesian kindwith the following theoretical issue: To what extent is it possible to devise a non-Brentanian property of intentionality, particularly one that is fully dissociated fromthe property of representation? This issue is shown to be much deeper and more dif-ficult than it looks once the nature of representation is properly apprehended; itseems to be still waiting for an answer in the current search for an alternative modelof the mind, if only because it has not yet be set in fully adequate terms.

1. ANTI-CARTESIANISM AND NEO-BRENTANISM: THEPROBLEM OF COMPATIBILITY

The search for alternative models of the mind has been going on for sometime now, but it remains arguably inconclusive. The assessment of what ithas achieved so far very much depends, however, on the way it is to be

Jean-Michel Roy has a main affiliation with �Ecole Normale Sup�erieure of Lyon (Lyon) asProfessor of Philosophy, France, and a secondary affiliation with East China Normal Uni-versity where he is a recurrent visiting Professor in the Philosophy Department.

The Southern Journal of Philosophy, Volume 53, Spindel Supplement (2015), 90–125.ISSN 0038-4283, online ISSN 2041-6962. DOI: 10.1111/sjp.12125

90

The Southern Journal of PhilosophyVolume 53, Spindel Supplement2015

understood, and serious uncertainty remains in this regard. As a matter offact, the notion itself of a search for alternative models of the mind standsin need of clarification on at least two basic and related points: Whatattempts to provide an understanding of the mind does it really refer to?And, to what view of the mind, exactly, are such attempts supposed to offeralternatives?

One natural answer to these two questions is to assimilate this search withthe series of efforts to elaborate an alternative to the cognitivist conception ofthe mind that lay at the core of the “cognitive revolution” and, accordingly,to locate its starting point in the emergence of the multifarious criticisms ofcognitivism that followed. This answer is, however, disputable on severalcounts. One reason is that some alternatives to the cognitivist model of themind, such as the connexionist one, have been present from the inception ofthe cognitive revolution and are usually not included in the field of referenceof the notion, which is standardly understood with a more restricted tempo-ral extension. Another reason is that the cognitivist model is clearly viewed inthis search as a recent incarnation of a more general Cartesian view of themind that constitutes its real critical target. This, however, suggests that thesearch should then be understood as comprising at least all the early 20thcentury attempts to overcome the Cartesian tradition in the area of mindtheorizing—such as those of Dewey, Wittgenstein, Ryle, and Hiedegger—aproposal sinning even more by temporal overextension with respect to theway the expression is standardly used. Early 20th century anti-Cartesianismis seen at best as an ancestor to the current search, not as one of itsprotagonists.

A solution to these difficulties consists in understanding the search asincluding only the series of efforts to denounce and remedy the insufficien-cies of the neurocognitive re-orientation that the cognitivist enterprise, bornin the 1950s, took at the turn of the 1990s under the inspiration of theoristssuch as Patricia and Paul Churchland, Steve Kosslyn, and Michael Gazza-niga.1 One must also see, accordingly, the key theoretical issue lying at itscore as the following one: how much, and how, should cognitive science gobeyond the neurocognitive model of the mind? Interpreted in this morelimited way, the search for alternative models of the mind is not discon-nected from the rejection of cognitivism and of cognitive Cartesianism, butonly represents a specific and recent stage in it. As a matter of fact, one liet-motifs in this search is the accusation that the break with cognitivism, andmore generally with cognitive Cartesianism, incarnated by the neuro-cognitive turn lacks radicality.

1 See Roy 2001.

91THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

Though more restricted with this perspective, the search for alternativemodels of the mind still covers a rather bewildering variety of attempts totransform the principles of cognitive explanation. Enactive, extended,embedded, situated, embodied, dynamical, evolutionary, phenomenological,constructivist, pragmatist, etc., alternative models of the mind are, in fact, anembarrassment of riches—an embarrassment raising a twofold importantcritical issue. The first issue is an interpretive one that consists in clarifyingthe specific content of each model and its relations of similarity, difference,and compatibility with the others; the second is an evaluative one that con-sists in arbitrating between the models. As a matter of fact, several attemptshave already been made in this direction.2 And although these attempts pro-vide different views of the situation, an interesting point of consensus emergesfrom them: namely, the idea that the search for models of the mind alterna-tive to the neurocognitive one finds a minimal theoretical unity in the factthat it is fundamentally one for a non-Cartesian conception of the mind.3

This minimal consensus is sufficient to raise an important issue regarding thesoundness of its general perspective.

The issue arises from the additional fact that this basic anti-Cartesianismarguably goes hand in hand with a persisting neo-Brentanism, inheritedfrom the early cognitivist steps of cognitive science. Indeed, the prevailingversion of cognitivism committed cognitive science not only to a mentalistform of explanation, that is to say to one endorsing the explanatory relevanceof mental properties, but also to a certain form of Brentano’s thesis, accord-ing to which intentionality is one of the marks of the mental, as well as a natu-

ralizable mark. This explicitly neo-Brentanian claim, pervasive among thefirst generation of philosophical analysts of cognitive science, seems to havebeen essentially preserved, although perhaps in a more discrete way,throughout the several waves of criticism cognitive science went through,including in the specific one under examination, which hardly offers anyrejection of neo-Brentanism and even contains some strong reaffirmations ofit. The difficulty, however, is whether this lingering neo-Brentanism is com-patible with anti-Cartesianism. Consequently, the issue is whether cognitivescience can consistently search for an alternative to the neurocognitive modelof the mind reached in the early 1990s that is both anti-Cartesian and neo-Brentanian. This issue of compatibility has largely gone unnoticed. It is, how-ever, of crucial importance for the receivability of the current search for,what Rowlands aptly labelled “a new science of the mind.”4

2 See, in particular, Wheeler 2005 and Rowlands 2010.3 See Roy 2011.4 See Rowlands 2010.

92 JEAN-MICHEL ROY

But why might there be an issue of compatibility in the first place? Theanswer, in a nutshell, is that neo-Brentanism has all the appearances of aCartesian kind of conception of the mind. Of course, this assertion needsjustification, ideally based on a rigorous definition of cognitive Cartesian-ism such as those proposed in the literature. In order to justify it, how-ever, it is enough to make the minimal assumption that neo-Brentanismshares at least one important theoretical feature with Cartesianism. A rea-sonable claim in this regard is to assume what I propose to call the thesis

of the representational essence of the neo-Brentanian property of intentionality and willformulate as follows: neo-Brentanism shares with cognitive Cartesianismthe essential feature of representationalism through the introduction of anessential link between the properties of intentionality and representation.Further, it is enough to make the additional and even less controversialassumption that anti-representationalism is a key feature of cognitive anti-Cartesianism to legitimize the suspicion that it looks inconsistent to searchfor an alternative to the neurocognitive model of the mind that is bothantirepresentationalist and neo-Brentanian. This assumption, then, legiti-mizes the issue of compatibility under a limited form that I propose tocall, in turn, the problem of anti-representationalist neo-Brentanian intentionalism

and will examine in the following pages.Put in its full form, this problem divides into three essential questions. The

first one expresses its purely theoretical dimension and can be phrased as fol-lows: (Q1) Can any model of the mind, in principle, combine the anti-representationalism of anti-Cartesianism with the intentionalism of neo-Brentanism, given the essential link existing between the neo-Brentaniannotion of intentionality and the notion of representation? The other twoexpress the interpretive and the evaluative sides of its critical dimension,respectively, and can be put in the following terms: (Q2) In its search for analternative to the neurocognitive model of the mind, does cognitive scienceactually try to combine these two elements, and, if so, how exactly? (Q3)How successful are such attempts?

In a strategy of divide and conquer, the ambition is here limited to dealingwith the first one of these three questions, proceeding in two successive steps.The first step consists in providing a negative answer, through a defense ofthe thesis of the representational essence of neo-Brentanian intentionality;the second step explores some of the implications of this incompatibilistanswer. Indeed, it is arguable that one important and immediate conse-quence is to confront any search for an alternative to the neurocognitivemodel of the mind with a crucial dilemma: give up either on radical anti-representationalism or on neo-Brentanism. One particularly interesting wayout of this dilemma consists in devising a non-Brentanian intentionalism.

93THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

This way out is, nevertheless, quite challenging, as it requires addressing thefollowing additional and quite difficult theoretical question: Is it possible,and if so how, to relax the deep connection established by neo-Brentanismbetween the property of intentionality and that of representation? The inves-tigation of this question leads directly to the critical dimension of the prob-lem of anti-representationalist intentionalism (Q2 and Q3 above), becausethe best strategy in order to come to terms with this theoretical difficultyseems to be to examine whether any of the alternatives to the neurocognitivemodel of the mind currently under development contains an acceptable solu-tion to it.

2. THE NEO-BRENTANIAN PROPERTY OF INTENTIONALITY

The key to the problem of compatibility lies essentially in the validity of theclaim that the neo-Brentanian property of intentionality entertains an intrin-sic link with the property of representation, as this claim is the most disputa-ble of all the ones on which it stands. In order to determine whether and inwhat sense it is acceptable, it is, however, no less essential to examine firstlyhow each of these two properties is to be properly understood.

The view of intentionality pervasive in the contemporary literature makesit essentially a distinctive property of mental states mainly characterized interms of aboutness, understood as the fact of being about something—usu-ally different from the state itself except in the special case of reflective inten-tionality—and, hence, a form of relational property. The notion ofaboutness is itself mostly left unanalyzed, when not explicitly declared unde-finable, although it is frequently parapharased with the help of a set of recur-ring notions such as contentfulness (an intentional state is a state with acontent), directedness (an intentional state is a state directed at something),reference (an intentional state is a state referring to something), attitude (anintentional state is an attitudinal state), or satisfiablity (an intentional state isa state with conditions of satisfaction). The notion of something hardly bene-fits from more analytic efforts but is, under the frequent labels of intentionalobject or content borrowed from the intentionalist tradition, intuitivelyassimilated with a quite general ontological category of objective entity. So,the property of intentionality remains on the whole, in the contemporary lit-erature, rather poorly defined in the intentional idiom itself. This is a prob-lematic situation in many respects that cannot but reflect a theoreticalembarrassment. As a matter of fact, as John Haugeland correctly remarked,“Intentionality is hard to get a glove on.”5

5 Haugeland 1990, p. 383.

94 JEAN-MICHEL ROY

The contemprary property of intentionality, for the most part, is explicitlyreferred back to Brentano’s 1874 Psychology from an Empircal Standpoint,6 and isintended as a neo-Brentanian property differing from the original view ontwo main points: it is seen neither as the unique mark of the mental, nor asone incompatible with a naturalist perspective. Despite such differences, thisBrentanian rooting recommends going back to Brentano’s view itself in orderto shed more light on the conception of intentionality at work in the searchfor alternative models of the mind. What was Brentano’s basic insight whenhe decided to give a new life to this scholastic notion?

The central source of Brentano’s introduction of the notion is a remark-ably ambiguous text: “Every psychical phenomenon,” Brentano writes, “ischaracterized by what the scholastics of the middle ages called the intentional(also the mental) inexistence of an object, and what we might call, thoughnot wholly unambiguously, relation to a content, direction towards an object(which is not to be understood as a reality), or immanent objectivity. Eachcontains something in itself as an object, though not each in the same way.”7

What this text essentially states is a hypothesis about the structure of men-tal states, which can be summarized as the claim that a mental state is char-acterized by a directional relation to an object. However, four differentdeterminations can be distinguished in this definition, raising an interroga-tion as to which of them corresponds to the essential feature of intentionality.Is it: (a) the fact that the object to which there is a relation is an immanentone, and as such something contained in the state (i.e., a content)? Or (b), thefact that the transcendant correlate of this immanent object is dispensable?Or (c), the directional character of the relation to the object, whatever thisreally means? Or, finally, (d) the very fact that the relation is one to some-thing as an object, or what might be called a relation of objectivation, in thesense of a relation in which the relatum is made into an object by contrast torelations in which it is a mere thing?

Some of these ambiguities have not gone unnoticed among readers of theBrentano thesis, and they lie at the source of the protracted debate of inter-pretation it has generated. Spiegelberg has, for instance, opted for the firstpossibility,8 and Roderick Chisholm for the second one.9 I personally wish toadvocate the view10 that, in 1874, all four determinations contribute to theessence of intentionality in Brentano’s eyes, although the most fundamental

6 See Brentano 1874.7 Brentano 1874, Book II chap 1, §5, p. 67 (translation amended).8 See Spiegelberg 1970.9 See Chisholm 1952.10 A slightly reformulated version of the one previously elaborated in Roy 2010, Chapter

IV, “Les theses de Brentano.”

95THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

one is the feature of objectivation. According to this interpretation, Brenta-no’s driving intuition in re-introducing the notion of intentionality is that thekey distinctive character of mental states is that they make a world of objects,as opposed to a world of things, emerge and, consequently, make the sub-ject/object opposition possible. Only creatures endowed with mentalityrelate to the world as subjects to objects, and this relation of objectivation isalso one of directedness, to an object which is immanent and has a dispensa-ble transcendant correlate. As already suggested, it is not clear, however,what the feature of directedness means or what it adds to the notion of objec-tivation itself. It nevertheless definitely points toward an idea of characteriza-tion on the part of the subject pole of the intentional relation. Accordingly, Ipropose to integrate it with the feature of objectivation and to further definethe core of Brentanian objectivation as the fact of not only being related tosomething as an object, but more specifically of characterizing something asan object. In this view, Brentanian intentionality is fundamentally thecapacity to characterize the world as an objective entity, and only creaturesendowed with mentality possess this capacity. In a way, it is a view ratherclose to the fairly standard notion that Brentanian intentionality ultimatelyreduces to that of objective reference, an opinion to be found even in Spei-gelberg’s and Chisholm’s interpretations in relation to their analyses of theevolution of Brentano’s thesis. Nevertheless, it differs from the standard viewby giving to the notion of objective reference the much more radical, andsomehow constitutive, sense of being the source of objectivity as such.

Several facts speak in favor of such an interpretation. The first one is theinsistence of Brentano himself on the importance of the feature (d) of objecti-vation. It is the only one he mentions when he summarizes his views in §9 ofChapter V of the 1874 book and writes: “We have found, as a distinctive fea-ture of all psychical phenomena, the intentional presence, the relation tosomething qua object [under the title of object]. No physical phenomenonoffers something of the kind.”11 The second reason is that it corresponds tothe way Brentano’s thesis was understood by some of his most distinguishedreaders and advocates, notably by Bertrand Russell, who until 1918 was anexplicit, although little known and nonorthodox, adopter of the thesis.12

Indeed, this adoption consisted fundamentally in assimilating his notion ofacquaintance, for him the building block of all cognitive mental phenomena,with Brentano’s notion of intentionality. Acquaintance was for Russell pre-cisely the relation through which the “fundamental fact” of the “dualism of

11 Brentano 1874, p. 74 (translation amended).12 See Roy 2010, Chapter VII, for further details.

96 JEAN-MICHEL ROY

subject and object”13 as such gets established,14 and whose converse is therelation of presentation.15 Through this assimilation, he thus unambiguouslymanifested that he understood Brentanian intentionality as a relation ofobjectivation of type (d), but one that rejects the notion of immanent objector content and retains solely that of transcendant object. Finally, the evolu-tion of Brentano provides further justification, since Brentano himself latergave up the notion of immanent object and only retained that of transcen-dant object, making determinations (a) and (b) in the end irrelevant and dem-onstrating retrospectively that determinations (c) and (d) were the key ones.

If objectivation as defined above lies at its core, the Brentanian notion ofintentionality is more complex and involves a distinction between the charac-ter of objectivation itself and its various modalities, as well as one between aprimary and a secondary, or reflexive, objectivation. Having no immediatebearing on the issue under scrutiny, these complexities of Brentano’s analysiswill nevertheless be left aside here.

3. THE NATURE OF REPRESENTATION: THE STAND INCONCEPTION

Turning to the property of representation, one is faced with a parallel feelingof elusiveness as well as of an underlying theoretical difficulty in the contem-porary cognitive literature—a feeling rather well captured by Rick Grush,who observes: “As crucial as this notion is, and as much theoretical attentionand industry has been devoted to it, it remains a frustrating enigma.”16

Elusive as it remains, a clearly identifiable general conception of its nature,that I propose to call the “stand in conception,” nevertheless prevails, in myopinion, in both the neurocognitive model of the mind and its antecedentsand in the search for alternative ones. According to this conception, a repre-sentation is basically something that stands in for something else. Its pervasivenesscan be illustrated by numerous references, some of them more explicit thanothers. Given the importance of this point, extensive quotation is in order.William Bechtel writes, for instance:

The ‘term’ representation is used in a variety of ways in cognitive science, makingit challenging to assess the different claims scientists make about representations . . .it is useful to begin with by distinguishing two aspects of representations, the func-tion of a representation as standing in for something else, and the format employed inthe representation. . . . A major strategy in cognitive science has been to explain

13 Russell 1918, p. 153.14 Russell 1913, p. 35.15 Russell 1918, p. 152.16 Grush, 1997, p. 349.

97THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

how an organism is successful in negotiating its environment by construing some ofits internal states or processes as carrying information about, or standing in for, thoseaspects of its body and external states or events that it takes account of in negotiat-ing its environment.17

Similarly, John Haugeland offers the following definition:

if the relevant features are reliably present and manifest to the system . . . wheneverthe adjustments must be made, then they need not be represented. Thus plantsthat track the sun with their leaves need not represent it or its position, because thetracking can be guided directly by the sun itself. But if the relevant features are notalways present, then they can, at least in some cases, be represented; that is, some-thing else can stand in for them, with the power to guide behavior in their stead.That which stands in for something else in this way is a representation; that which itstands in for is its content; and its standing in for that content is representing it.18

Tim van Gelder is also particularly worth quoting in this regard: “. . . anyreasonable characterization [of the notion of representation is] based aroundthe core idea of some state of a system which, by virtue of some general rep-resentational scheme, stands in for some further state of affairs, thereby ena-bling the system to behave appropriately with respect to that state ofaffairs.”19 Finally, Rick Grush asserts on his side: “Representations are enti-ties which stand for something else – or better they are entities which are usedto stand for something else.”20

In virtue of this definition, a representation is thus something acting as asubstitute for something else, an entity that is present in replacement of some-thing else. This view corresponds quite strictly to the most traditional definitionof the symbol as an aliquid quod stat pro aliquo and deserves, accordingly, to beconsidered as a symbolic conception of representation. In this perspective,then, a mental representation is nothing else than a symbol of a mental kind,whatever this specifically mental dimension really comes down to. There is lit-tle doubt that partisans of the pervasive representational theory of the mindshare such a perspective, as their occasional swtiching from the notion of repre-sentation to that of symbol, well illustrated by Fodor21 confirms.

Beyond this basic characterization in terms of symbols, the stand in concep-tion is typically further developed with the help of a set of recurrent notions,including those of reference, information, function, and satisfiability. The result-ing standard conception of representation is that of an information bearing element

17 Bechtel 1998, p.3.18 Haugeland 1998, p. 172.19 van Gelder 1995, p. 351.20 Grush 1997, p. 349.21 See, for instance, Fodor 1987.

98 JEAN-MICHEL ROY

functioning as an information provider about something else than itself that constitutesboth its referent and its satisfier. This definition is not fundamentally wrong, but itlacks precision, rigor, and clarity in several respects, and, thereby, it incurs therisk of confusing a specific kind of representation with the genus itself. This, inturn, affects the solutions that can be brought to a host of crucial issues hingingon the understanding of the nature of representation that range from its elimi-nability to its naturalizability, as well as, precisely, to its relation with the prop-erty of intentionality. It is not the least of paradoxes of contemporary cognitiveliterature that is neglects the problem of the nature of representation in favor ofother such issues, although their solution depends crucially on the way thisproblem is solved. Consequently, what is needed is a full blown theory of thenature of representation that carefully lays out its most essential and basic fea-tures and rigorously identifies its key paramaters of specification. Only on thebasis of such a theory can its relations with the notion of intentionality beadequately grasped.

4. AN ALTERNATIVE ANALYSIS OF THE NATURE OFREPRESENTATION

My ambition here is not to provide such a theory, but the grounds of analternative view of the principles on which to erect one, with a sufficientdegree of development for establishing the thesis of the representationalessence of the neo-Brentanian property of intentionality and, therefore, thereality of the problem of anti-representationalist neo-Brentanian intentional-ism. This alternative view is centered on the following three essential claims:18) the faculty of representation is best understood at its most basic level as afaculty of specification, i.e., a capacity to specify things in the sense of attrib-uting them determinations; 28) among the possible forms such a specificationmight assume, a key distinction must be introduced between a direct and anindirect one; 38) the direct one is the most basic. The strategy to be followedfor establishing these three claims consists first in reformulating the stand inconception, and then arguing why and how one must go beyond it in orderto appropriatley capture the essential nature of representation.

4.1. Reformulating the Stand In Conception

Contrary to what Nelson Goodman seems to think,22 but in agreement withHusserl,23 cases of external representations are not degraded forms of

22 Goodman 1976, p.5.23 See Husserl 1900–1901.

99THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

representing. They even constitute the best candidates for correctly under-standing the stand in conception, because they are the ones that best fit it.

What do we mean when we claim, for instance, that an ambassador repre-sents a government? Taking the historical case of the first U.S. ambassadorto France as an example, it is clear that we mean that Benjamin Franklin isnot a member of that government, but a representative of it, in the sense ofan element that stands in a certain relation of representing with another oneplaying the function of represented element, in opposition to that of representing ele-

ment. The first key question, therefore, is to characterize this relation itself.Three essential aspects must arguably be distinguished in it.

It is in the first place a relation of substitution, which corresponds to thestanding in feature of the stand in conception. Indeed, qua representing ele-ment, Benjamin Franklin is what interacts with the French government inlieu of the U.S. government. He is what stands in front of that governmentwhen it addresses the U.S. government. He replaces it, acting as a sort ofersatz of it.

However, he is not a mere substitute of it, because a mere substitute justtakes the place of another element and somehow puts it out of the game. Onthe contrary, by interacting with Franklin, France is in fact dealing with theU.S. government itself. It is dealing with it through Franklin, that is to say,indirectly instead of directly. Accordingly, Franklin is not only a substitute of

the U.S. governement, but a substitute for the U.S. government. He stands inFrance for the U.S. government. This second relation, corresponding to the“for” element of the stand in conception, also corresponds fairly well to a cer-tain notion of reference, intuitively understood as the fact of one thing bringing

in another.Finally, it appears difficult that something could stand in for something

else without standing as such to some other thing and, therefore, withoutbeing apprehended as such by that thing. If there were nobody to whom theambassador were taken as a substitute of his government, he would not reallyqualify as an ambassador, at best as a merely potential one. Consequently, itlooks necessary to introduce in the analysis a complementary relation ofapprehension, as well as a complementary element acting as the subject of thatrelation of apprehension that will be labelled the apprehending entity. So, a rep-resentation is to be defined as an entity that, on the one hand, substitutes forand refers to some other entity and, on the other hand, is apprehended assuch by yet another one.

The represented element, which is taken here in the broadest possiblesense of whatever is the related term of the representing relation and, accord-ingly, as equivalent to what is now more usually designated with the overlyambiguous expression of “content of a representation,” has received more

100 JEAN-MICHEL ROY

attention than this representing relation itself. Numerous and often quitesophisticated distinctions have been introduced in the course of the history ofthe multifarious analyses of the nature of representation. Although they lackconsistency across theories, three of them are of particular importance andcan be presented as follows.

First, the notion of represented refers to the element of which there is a repre-sentation, that is to say, the element that exists independently of the relationof representation but happens to enter into it in the specific position ofrelated term. Like, for example, the group of people constituting the U.S.government who happens to enter, among many other kinds of relation thatit entertains, into a relation of diplomatic representation, in the position ofwhat Franklin is a substitute of and refers to. It is best to label this elementthe represented entity. Even though it implies in some sense the existence of thisentity, this first distinction is free of any specific commitment as to its onto-logical status in the sense that it leaves all doors open to the full spectrum ofpossible ontological positions, from hard realism to irrealism, and thereby toa variety of corresponding forms of representationalism ranging from realistto irrealist ones.

Second, there is the represented as represented or as such, that is to say, asrelated term intrinsically linked to the relation of representation. In this sec-ond sense, the represented designates what is represented in a representationeven if there is nothing of which it is a representation, as in the widely dis-cussed cases of “representations without objects” at the turn of the twentiethcentury. For example, what Franklin would represent as a U.S. ambassador,even if the U.S. were, in fact, a sheer fiction because the independence fromthe Great Britain had not been acheived. To emphasize its intrinsic connec-tion with the representing relation, it is best designated as the term of therepresenting relation or the representational term. This second distinction cer-tainly is the most debated one in the literature on the nature of representa-tion, because it raises numerous and quite difficult issues. One such issueconcerns its ontological independence: How much should it be conceived asan entity of its own, separate from the represented entity? If it is not, howcan it survive the disappearance of the represented entity and not turn a rep-resentation without an object into a nonrepresentation? If it is, what kind ofentity is it really, and how does it relate to the represented entity? Possibleanswers to these questions correspond to as many possible specifications ofthe notion of representation, in addition to those opened by the degree ofreality to be granted to the represented entity—quite a number of whichhave been to some extent explored in the past.

Third, what is represented might correspond more particularly to the spe-cific aspect under which the representational term figures in the

101THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

representational relation, that is, the specific determinations with which it isrepresented. In other words, the way it is described. This third dimension ofthe represented is familiar because of its equivalent with the notion of sensein the analysis of linguistic meaning. As a matter of fact, some analyses donot hesitate to extend to it this term “sense,” such as the Husserlian theoryand its famous concept of Noematic Sinn. It is also what the expression of con-tent introduced in the Brentanian tradition by Twardowski24 and H€ofler25

was specifically designed to capture. Both the expression of sense and theexpression of content nevertheless raise difficulties: the first because it favorsan all too problematic tendency to analyze mental properties on the modelof linguistic ones, and the second because of the generic use that the term“content” has now acquired, making it designate the represented in itsentirety. For these reasons, it seems more appropriate to label this thirddimension of the represented the descriptive element. This further distinction hastraditionally brought additional complications to the issues elicited by theadmission of the representational term, raising, in particular, the questionwhether it should be considered as a separate entity and, if so, how it relatesto the other constituents of the represented.

This overall analysis of the basic nature of representation can be summar-ized with the different case of a photographic representation, like the famousLife magazine photography of Marilyn Monroe lifting her skirt and reveilingher legs. In this instance, the representing element is the material dimensionof the photograph, i.e., the Life magazine printed page. The representedentity is the individual Marylin Monroe, who happens to get involved in thisspecific relation of representation, but is also involved in many other ones.The representational term is the individual Marilyn Monroe as representedby the Life magazine page, who would still qualify as what this page repre-sents even if Marilyn were a sheer fiction. The descriptive element corre-sponds to the specific way Marilyn Monroe appears on this picture, that is,with a certain posture, a certain dress. In addition, the printed page clearlyacts as a visual substitute of the individual Marilyn Monroe: it appears in themagazine in order to replace the actual Marilyn, that is, what I am supposedto see when I open this magazine instead of seeing her. It is not a mere sub-stitute of her, however, as I am also supposed, when I look at it, to have acertain form of visual experience of Marilyn and, hence, to be referred toher. It somehow brings her in. Finally, the apprehension element lies in thefact that I visually perceive the printed page in a certain way; this is in con-trast with a visual experience that does not identify it as a picture and reduces

24 See Twardowski 1894.25 See H€ofler 1890.

102 JEAN-MICHEL ROY

it to a patchwork of meaningless shapes and grey hues. Moreover, it is argu-able that this relation of apprehension is here again necessary, not only forrecognizing what it is, but for making it what it is.

It should be noted that in this analysis of external representation along thelines of a refomulated stand in conception, the apprehending relation is,nevertheless, made less a constituent of the representation than an indispen-sable complement to it. Even though it must be apprehended as such, whatreally makes a representation what it is, is the fact that it is the referring sub-stitute of something else. As a matter of fact, if one can assert that the Frenchgovernment apprehends the ambassador as a representative of the U.S. gov-ernment, it is incorrect to say that the French government represents (toitself) the U.S. government by means of its ambassador. It might be objected,however, that this limitation is due to the specific kind of external representa-tion taken into consideration and cannot be generalized. In the case of thepicture of Marilyn Monroe, it is legitimate to say not only that whoever seesthe printed page correctly apprehends it as a photographic representation ofMarilyn Monroe, but also visually represents Marilyn Monroe to himself.Shouldn’t the relation of apprehension be consequently integrated into thedefinition of the representation relation itself, instead of being seen as a mereadd-on to it, even if of a necessary kind? I favor a negative answer, but thequestion is a difficult one. For the sake of simplification I will leave it asidehere, as nothing essential for the demonstration at stake hinges on thisdecision.

How exactly, then, to extend this reformulation of the stand in conceptionof representation extracted from the case of external representations to thatof internal representations? It is necessary to first clarify how the frontierbetween the external and the internal should itself be drawn. The most natu-ral proposal, in this regard, is to make it coincide with the limits of what hasbeen labelled the apprehending element. That is, a representation counts asan external one when the representing element is external to the elementthat apprehends it as such, and internal when the representing element isinternal to the element that apprehends it as such. In other words, an inter-nal representation is an element Y internal to an element X such that Y isapprehended by X as standing in for an element Z, where standing in meansbeing a substitute of, as well as referring to Z. The key to understanding theinternalization of the above reformulation of the stand in conception is thusto understand how the representing element itself can be internalized. Thereseem to be two main ways of doing this.

The first one consists in finding an equivalent to what we might call thematerial entity that plays the role of representing element in the case of anexternal representation such as a photograph. In the perspective of cognitive

103THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

naturalism that has dominated the cognitive science enterprise from itsinception, some form of brain configuration, or brain and body configura-tion, is the obvious candidate. A solution well illustrated by Fodor’s versionof the representational theory of the mind, that fits neatly into the proposedreformulation. Indeed, Fodor explicitly assimilates a mental representationto a mental symbol M occurring in a cognitive system and standing in twocausal relations26: one with the rest of the system and corresponding at a cer-tain level with the relation of apprehension, and one of aboutness with anelement in the environment and corresponding to the substitution and refer-ence relation. In a natural cognitive system, this mental symbol M is realizedor implemented by some neuronal firing pattern, or even some specific neu-ral configuration, just like the photograph of Marilyn Monroe is realized invarious material structures from which it can also be abstracted.

There is, however, a different way of internalizing the representing ele-ment, which played a major role in the representationalist tradition. It doesnot simply assimilate the representing element to an internal equivalent ofthe material realizer of the external representation, but, rather, somehow dis-places it with respect to the architecture of external representation. Indeed,what is treated as the representing element is one of the first two elementsdistinguished in the represented in the case of the external representation,namely the representational term or the descriptive element, and they aresaid to represent what was labelled the represented entity. When switchingfrom the photograph of Marilyn Monroe to a mental image of her, what isrepresenting is the sort of duplication of her that we elaborate when imagin-ing and have in front of “the mental eye,” and not the material structure ofthis duplicated Marilyn, be it made of some specific mental stuff or of neuro-biological matter. This alternative form of internalization is made quite clearby the way Bertrand Russell, in his first period, summarizes it in order to putit into question. He talks about it as “the theory that between subject andobject there is a third entity, the content, which is mental, and is that thoughtor state of mind by means of which the subject apprehends the object.”27 Healso correctly makes clear the link uniting this view formulated in the Brenta-nian vocabulary of content, obviously borrowed from Alexius Meinong whowas his entry door into Brentanism,28 and the traditional theory of ideas,described in turn as “the view . . . that there is some mental existent whichmay be called the ‘idea’ of something outside the mind of the person who

26 See, in particular, Fodor 1987b.27 Russell 1913, p. 5.28 Meinong, who was also responsible for the introduction of the notion of content as he

co-authored H€ofler’s Logic, clearly saw it as a tertium quid that acts as a mediator in the inten-tional relation, as a sort of intentional relay.

104 JEAN-MICHEL ROY

has the idea . . . [and that] ideas become a veil between us and outside things. . . .”29 The framework is clearly that of the stand in conception with contentplaying the role of what substitutes for and refers to an object. And thisunderstanding of internal representation undoubtedly continues to dominatethe interpretation of cognitive representationalism, at least on the side of itsdetractors. In the same vein, Alva No€e speaks for instance of a view that seesrepresentation as an “internal-world model,”30 Mark Rowlands of a viewthat sees it as an “internal reproduction,”31 and Rodney Brooks as a“model.”32

4.2. Going Beyond the Stand In Conception (1): Dissociating Substitution from

Reference

What preceeds is no more than a skech of the basic principles of a reformula-tion of the stand in conception, calling for a full theoretical elaboration andfor a detailed critical confrontation with the main theories of the nature ofrepresentation that have been offered. It is sufficient, however, to put intolight one fundamental error in it. This error consists in confusing what is nomore than a specific type of representation, namely the indirect or symbolicone where something is represented by means of an intermediary, with therepresentational genus itself. The source of this confusion lies in the lack ofdistinction between the relation of substitution and that of reference, as ifthey were one and the same, or at least indissociable from each other. Thatis to say, as if it were impossible to refer to something without being a substi-tute of that thing. The way some representatives of the stand in conceptionexpress themselves is quite revealing in this respect. William Bechtel talks, forinstance, of “carrying information about, and so standing in for it,”33 whileVan Gelder uses the expression of “symbolic representation.”34 In a similarvein, Nelson Goodman writes that “the plain fact is that a picture, to repre-sent an object, must be a symbol for it, stand for it, refer to it.”35

However, once a clear distinction is introduced between the fact of beingthe substitute of something and the fact of referring to something, it appearsthat that there is no reason why something could not refer without acting asa substitute. Further, there is also no reason to deny that something thatrefers in a nonsymbolic or indirect way is representing. Moreover, because of

29 Russell 1918, p. 156.30 No€e, 2004, pp. 22–23.31 Rowlands 2010, p. 33.32 Brooks 1991, passim.33 Bechtel 1998, p. 296.34 van Gelder 1996, p. 351.35 Goodman 1976, p. 5.

105THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

its simplicity, it is also naturally suggested that nonsymbolic reference evenconstitutes the most basic and general constitutive feature of representation,and that representing symbolically is no more than one possible specificationof this basic feature obtained by adding the relation of substitution to that ofreference. According to this alternative characterization, and when consid-ered at its most general level, a representation is thus an entity Y that refersto an entity Z directly, in the sense that Y does not act in addition as a substi-tute of Z. No modification in the analysis of the represented is required.However, the complementary relation of apprehension is no longer necessaryat this level, as it is arguably required by the substitutional component of therepresenting relation. Although, one might still defend a more specific ver-sion of the view that sees reference as involving some kind of relationbetween the representing element and a subject or a user, in the spirit offamiliar user theories of representation. The definition is intended to remainneutral on the necessity of this relation, which is different from the onerequired by the property of substitution and that is to be eliminated with thatproperty itself.

Accordingly, it is indeed possible to treat the verbs “to represent” and “torefer” as synonymous, but not the verbs “to refer” and “to stand for.” Thislast one should be reserved for the symbolic type of representation. A termi-nological distinction already present in the literature can usefully be invokedfor clarifying and fixing the transformation advocated here in the analysis ofthe nature of representation. A number of authors, from some heirs of Bren-tano to a few contemporary theorists, have indeed proposed distinguishingbetween a presentation and a representation. Leaving aside for the momentthe specific content they give to this distinction, although it is not unrelatedwith the one advocated here, it could be said that the alternative claim putforward is that a representation is at its most basic level a presentation, andthat a representation properly speaking is nothing else than a more complexand specific type of presentation, characterized by its indirectness or substitu-tional character. However recommendable this claim is, such a terminologi-cal modification seems to go against too long of an established tradition to bereally useful. Accordingly, it seems preferable to keep talking in terms of rep-resentation, and to resort when necessary to an opposition between a presen-tational—and generic—representation and a symbolic—and specific—representation.

4.3. Going Beyond the Stand In Conception (2): Representation As Specification

By relegating the stand in conception to the status of specific definition,this alternative view also puts reference at the very heart of the general

106 JEAN-MICHEL ROY

nature of representation and, consequently, calls for an inquiry into theessential characteristics of reference itself in order to make further progressin the elucidation of this nature. This requirement does not make the taskany easier, since reference can be seen as one of the most basic notions ofthe whole philosophical tradition but also as one whose content, probablybecause it is so fundamental, has been left mostly unclarified, making it asort of philosophical blackspot. In recent times, views about its determina-tion at the language level or its naturalization at the mental level havecertainly been put forward, but these views say very little, if anything atall, about what reference is meant to be in the referential idiom itself. Acrucial and difficult task to be completed in order to shed full light on thenature of representation is therefore to provide an answer to the followingquestion: What do we mean exactly when we say that something refers tosomething else?

For all its vagueness, the rough and intuitive definition that has been usedso far, to the effect that it consists for the first thing to somehow bring in ormake present the second one, puts us on the right track toward the principlesof a more precise and technical definition. For the definition points to theidea that the notion of reference should be defined on the basis of that ofspecification, in the sense of a capacity of specifying or of attributing deter-minations, that is, as something that specifies something else as an object.With the consequence that the notion of specification is what captures, infact, the most general feature of the nature of representation, and also thatreference or what might also be called objective specification is only one par-ticular, although central, form of it. This twofold claim will be clarified andestablished in reverse order by going back to the examination of what isactually needed when a representation is postulated. Or, more exactly, whenit looks illegitimate to postulate one, as such negative cases are in fact morerevealing than positive ones.

Let us revert, for instance, to the classical situation of the causal explana-tion of the transmission of movement between two billiard balls that playedsuch a central role since it received Hume’s honors. The observed data—consisting in the simple fact that a first ball rolls towards another one, estab-lishes contact with it, and the second one then sets itself into motion—can beaccounted for in a variety of ways, such as the four following: 1) the first ballis driven by an internal force, and the impact transmits this force to the sec-ond one; 2) the first ball is driven by an internal force, and it so happens that,at the precise moment of contact with the second ball, this second ball isdrawn up by an external force of stronger magnitude and exerted in the verysame direction; 3) the first ball is driven by an internal force, and it so hap-pens that, at the precise moment of contact with the second ball, this second

107THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

ball jumps forward as a result of a sudden upsurge of an internal forceexerted in the very same direction; or 4) the second ball knows that the firstone is moving towards itself and therefore decides to run away from it at thevery moment when it gets in contact with it.

Although all equally account for the specific data to be explained, thereis a clear heterogeneity between the first three and the last one. The firstgroup is integrated by explanations of a purely mechanical kind, in thesense of explanations that only resort to forces and movements.The fourthone, on the contrary, gives explanatory relevance to such notions as know-ing and deciding that can be seen as representational ones in the sense ofpresupposing the notion of representation. It can accordingly be rephrasedas follows: the second ball entertains a representation of the first one asmoving toward itself on a collision course and takes this representation asa causal determinant of its physical movements. It is important to empha-size that nothing hinges on the validity of this representational reformula-tion of the notions of knowing and deciding; should one dispute it, it isthen enough to take the explanation couched in representational terms asthe original one. The difference, however, is not solely one of content butalso of acceptability. The representational explanation is clearly unaccept-able in this case, in a way and to a degree that none of the others is. Themain reasons for this are no less clear. It violates a principle of simplicity:much less complex explanations are available. It also goes against theprinciple of ontological economy and thus falls victim to Occam’s razor.Finally, it also sins by anthropomorphism, uselessly treating the secondbillard ball on the model of a human cognitive system and offering anaccount of a natural phenomenon that belongs to what Auguste Comteliked to call the theological age of humanity.

But what exactly is the essential idea that we should thus save by elimi-nating the explanation of the sequence of movements of the billards ballsin representational terms? It is undisputedly the twofold one that the sec-ond ball is endowed with a capacity to apprehend outside things in a cer-tain way, that is to say, to elaborate a characterization of its environment,to attribute a number of determinations to it, or, to put it in other wordsstill, to provide a certain specification of this environment and then usethis specification as a determinant of its behavior. The first of these twoelements is no less certainly the most crucial one, as a billard ball with thecapacity to specify its environment but not that of using this specificationas a determinant of its behavior would certainly count as a billard ballendowed with the capacity to represent. To put it in a nutshell, what therefusal to grant representations to the billard ball puts into light is thatthe world of representation emerges as soon as the very basic notion of

108 JEAN-MICHEL ROY

what there is according to or for something emerges, as opposed to whatthere is in itself. From this perspective, an explanation of the billard ball’smovement is fundamentally representational inasmuch as it accepts therelevance of the idea that something in the second ball characterizes itsenvironment as one in which the first ball is on a collision course towarditself.

Claiming that representing ultimately comes down to specifying, andconsequently that something is a representation inasmuch as it functions aswhat might be called a specifier, is not unrelated to its standard character-ization as information provider. But this view makes it something moreelementary—since not all specifying qualifies as an informing—and as aresult, something also that is not intrinsically epistemic. Furthermore, itmakes it something more elementary than referring, for the property ofreference can indeed be defined as a particular, even if central, form ofspecification, namely as an objective one in the sense of a characterizationof something as an objective entity, whatever this ultimately means. Itseems indeed reasonable to think that not all representation, so under-stood, deserves to be considered as a representation of an object. Thewell-known dotted pictures from which slowly emerges the perception of adalmatian dog provides, for instance, a good case where one can be seenas switching from a nonobjective representation to an objective one. Andin such a perspective, the representational term and the descriptive ele-ment of a representation are not two different kinds of properties of a rep-resentation but, rather, two different aspects of one and the samecharacteristic of specification: the sense corresponds to the way in whichsomething is specified, beyond being specified as an objective entity.

The double claim that a representation is fundamentally a presentationand that a presentation is fundamentally a specification results in anextremely minimalist form of representationalism, not in the standardsense of the notion of representational minimalism—that corresponds to aposition limiting the recourse to the property of representation—, but inthe sense of a theory of the nature of representation that puts very mini-mal requirements on qualifying as something representational. The conse-quence is that representationalism runs much deeper than is usuallyassumed in the debates about the possibility of devising an anti-representational alternative model of the mind, and that radically breakingaway with it definitely needs more than rejecting the postulation of ele-ments that stand in for something, Another of its consequences is that ref-erence is not as crucial to representation as the stand in conception tendsto have us believe. From this perspective, reference is only essential to aspecific type, however encompassing and central, of representations.

109THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

However, reference is thereby made on the contrary an essentially repre-sentational notion.

4.4. Converging Contemporary Views

Even though it differs from them, this minimalist representationalism is notwithout its echoes of some contemporary views. Despite the fact that he sideswith the stand in conception, the insistence with which Mark Rowlandsemphasizes that a representation is primarily something that “makes claimsabout the world”36 is, for instance, clearly consonant with its second compo-nent, namely the thesis that representation is essentially specification. Therecently developed radical enactivism of Daniel Hutto and Eric Myin is evenclearer in this respect. For in their effort to get rid of any remnant of thenotion of representational content in accounting for basic cognition, they areled to identify it in the very same terms used above: “To qualify as represen-tational, an inner state . . . must, so to speak, have the function of saying orindicating that things stand thus and so. . . .”37

Similarly, the presentational component of the definition finds counter-parts in authors like John Searle or Rick Grusch, who argue to a differentextent for a distinction between representation as such and representation aspresentation. In fact, Searle starts by defining representation at its most gen-eral level in a way that is extremely close to the specification view.38 Indeed,a representation is in his eyes anything that both has conditions of satisfactionand is associated with what he explicitly calls a “determination” of these con-ditions, in the sense of a certain specification of their nature under a certainaspect, and that he labels a representative content. This general characteriza-tion is free from any reference to a stand in element, as not even the repre-sentative content is analyzed as playing the role of an intermediary withrespect to the conditions of satisfaction. Furthermore, Searle introduces onthis basis a distinction between representations, so understood, that are pre-sentations, such as visual perception, and those that are not, such as beliefsand desires. But if the distinction has to do with an opposition between adirect and an indirect apprehension of the conditions of satisfaction, it is notthe one that opposes a nonsymbolic to a symbolic form of representationaccording to the above analyses. It corresponds, rather, to the Husserlian dis-tinction between a form of representation that specifies its object as presenteffectively, in itself, to one that does not. There is, however, in Searle a fur-ther distinction between presentation and representation that conforms

36 Rowlands 2010, passim.37 Hutto 2013, p. 62.38 See Searle 1983.

110 JEAN-MICHEL ROY

squarely with these analyses. In the particular area of visual perception,Searle introduces a further difference between a nonrepresentative theory ofvisual perception, that understands perception as a presentation in the sensenot only of a representation that delivers its object as directly present itselfbut of one whose structure also involves no stand in element, and a represen-tative theory view of perception that does not introduce such a stand in ele-ment. However, this opposition between nonsymbolic and symbolicpresentative kinds of representation only plays a secondary role.

Grush’s version of the distinction also arguably presents some importantanalogy with the one advocated here, since the distinguishing feature of a rep-resentation seems to lie, in his opinion, in the fact that it is “an entity [used] asan off-line stand-in”39 apprehended as a “model” of something else, while apresentation is something simply used as an information carrier about some-thing else without standing in as a model for it. Nevertheless, an important dif-ference between the two analyses, beyond the fact that Grush bases his own onthe standard notion of information carrying, is that he sees a representation ascarrying less information than a presentation, while no intrinsic difference ismade between a symbolic and indirect way of specifying and a direct one.

5. THE REPRESENTATIONAL ESSENCE OF NEO-BRENTANIANINTENTIONALITY

On the basis of these developments, the issue of the validity of the thesis ofthe representational essence of the neo-Brentanian property of intentionalitycan now be confronted on more precise grounds. Is it possible to claim thatthis property entertains an essential connection with that of representationand, in addition, that this essential connection makes neo-Brentanian inten-tionalism incompatible with anti-representationalism?

The answer is pretty straightforward. If correct, the previous analysescommand indeed an understanding of neo-Brentanian intentionality as anintrinsically representational property, although in the presentational senseof the notion of representation, in other words, as an intrinsically presenta-tional property. And the reason is simply that they reveal neo-Brentanianintentionality to be a specific and central form of representation, in virtue ofthe fact that the notion of objectivation that lies at its core is identical to thatof reference understood as an objective form of specification. Neo-Brentanian intentionality is essentially representational because it is essen-tially a specification of an objective kind.

39 See Grush 1997.

111THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

This analysis might be considered insufficient, however, because by result-ing in the sheer identification of the notion of intentionality with that of refer-ence, it deprives the former of any specificity, in contradiction with the deepintuition driving its introduction by Brentano and its preservation by mostcontemporary cognitive theorists. However, it is in fact preferable for a num-ber of reasons to enrich the analysis of representation as specification withthe introduction of a further distinction between reference tout court andobjective reference. This distinction is based on the difference between speci-fying something simply as an entity and specifying something as an object(whatever, once again, this comes down to). The objection raised can therebybe easily circumvented, as the identification must consequently be amendedand intentionality be made identical with objective reference only.

However, if the thesis, so understood, of the representational essenceof Brentanian intentionality is correct, then the problem of anti-representationalist neo-Brentanian intentionalism is also undisputably genu-ine. For accepting the relevance of the neo-Brentanian property of intention-ality is accepting that of a specific form of the property of representation.Neo-Brentanian intentionalism is not dissociable from representationalismcorrectly apprehended.

The theoretical considerations supporting this twofold conclusion can beaided by historical ones, for there is certainly nothing new in the thesis of therepresentational essence of the property of neo-Brentanian intentionality. Inthe first place, it corresponds on the contrary to the dominant interpretationof Brentano’s thesis in cognitivist and even neurocognitive times, whenbasically all theorists, friends or foes of the thesis—including people withtheoretical orientations as varied as Daniel Dennett, Paul and PatricaChurchland, John Searle and Jerry Fodor—analyzed it in representationalterms and made it a central tenet of cognitive representationalism. Further-more, this interpretation is largely in continuity with the one it had receivedantecedently, and it finds undeniable roots in the Brentanian schoool itself,and particularly in Brentano.

Indeed, on the one hand, Brentano clearly held a form of the thesis ofthe representational essence of intentionality by making representation themost basic intentional modality on which the other two that he accepted,judgment and feeling, are based. As a result, every mental state, as statedin a famous formula of the 1874 book that Husserl later put at the centerof his critique of Brentanism in Part IV of his Logical Investigations, either isa representation or is based upon a representation. As a matter of fact,intentionality is only one of several marks of the mental explored in the1874 book, and representation is one of these marks. Brentano did notreject its relevance in the least, but he considered intentionality to be a

112 JEAN-MICHEL ROY

better one. Furthermore, the representationalist structure of Brentanianintentionality is astonishingly close to the presentational one advocated inthe above theoretical developments. For what is the essential feature ofthe representational modality of intentionality according to Brentano? Theterseness of the answer he provides to this question is not proportional toits importance. He says enough, however, to show that he leans definitelyon the side of a sort of presentationalism. Representing is defined byBrentano as the sheer fact of presenting an object as such. “We talk ofrepresentation,” he writes for instance, “any time an object is shown tous.”40 Although it is not specified, the notion of showing used here lendsitself to an interpretation in terms of manifestation and, hence, of a directrelation with an object and not an indirect one.41

It can therefore be concluded that he not only held a thesis of the repre-sentational essence of intentionality, but under a form very similar to the onedeveloped here. One important difference is surely that representation wasfor him a modality of intentionality and not simply a structural feature of it.But this difference is a surface one, since the representational modality is onethat lies at the heart of every other modality. A more significant differencethat still distinguishes the two analyses, however, is the fact that for Brentanoall representation was an objective one, or, in other words, that there was nononintentional representation.

The case of Alexius Meinong, a major element of the Brentanian legacygiven its influence on the analytical current lying predominantly at the rootof the cognitive science enterprise, also a telling, although it is too complexto be dealt with in detail. First and foremost, Meinong made representa-tion, no less than Brentano did, the most basic form of intentionality. More-over, he introduced an explicit distinction between representation andpresentation. But presentation was for him only a constituent of the repre-sentational kind of intentionality, even if a determinant one, since it was theone with the objectivating role. However, the introduction of this presenta-tional element in the analysis of the basic representational modality ofintentionality did not prevent him from also giving it a symbolic structurethrough the introduction of the additional distinction between content andobject that, in his case, arguably makes the first one something standing in

40 Brentano 1874, p. 158.41 Additional considerations regarding the difficult question of the status of the intentional

object in his theory and based on Brentano’s correspondence with his former students (in par-ticular his famous letter to Anton Marty of March 15 1905 [see Brentano 1930]) converge,in my opinion, on the same conclusion. It is to be noted that the English version of the 1874book translates Vorstellung to presentation. Although an erroneous choice in strict terms of trans-lation, it is probably a correct one theoretically speaking.

113THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

for the second one.42 So, if Meinong’s intentionality is no less representa-tional than Brentano, it is so in the symbolic sense rather than the presenta-tional sense.

6. IMPLICATIONS: THE PROBLEM OF NON-BRENTANIANINTENTIONALISM

The reality of the problem of anti-representationalist neo-Brentanian intention-alism confronts alternative theories of the mind that pretend to combine thetwo incompatible claims lying at its core with an obvious dilemma: either torenounce anti-representationalism or to forego neo-Brentanian intentionalism.

It is to be noted that only anti-representationalism stricto sensu, or radicalanti-representationalism, is at stake in the first branch of this dilemma—thatis to say, the view that no theoretical relevance whatsoever should be grantedto a property of representation in the theory of the mind. For the option isstill left open to grant some relevance to this property in the proportiongranted to the property of neo-Brentanian intentionality, giving rise to amoderate form of representationalism that corresponds to what is currentlyoften designated as a minimal one. This moderation can alternatively beexercised not with respect to the extension of the property of representation,but to the forms of representation required by the acceptance of intentional-ism, opting, for instance, for a purely nonsymbolic representationalism if thisis what is needed for preserving neo-Brentanian intentionality.

Similar caution is in order concerning the second branch of the dilemma.Renouncing neo-Brentanian intentionalism means refusing to grant any the-oretical relevance to a neo-Brentanian property of intentionality and,accordingly, to opt for its elimination. But this anti-intentionalism or inten-tional eliminativism can be understood in a radical or in a moderate way aswell. The radical perspective is tantamount to rejecting intentionalism alto-gether under the assumption that there is no other possible property of inten-tionality than the neo-Brentanian one. Refusing such an assumption, themoderate perspective accepts that the relevance of a non-neo-Brentanianproperty of intentionality remains a theoretical possibility to be exploredand, consequently, that having to renounce neo-Brentanian intentionalismdoes not mean yet having to renounce intentionalism at large.

In contradistinction to the solutions consisting either in renouncing radicalanti-representationalism or opting for intentional eliminativism, whichboth bite the bullet of the incompatibilist implications of the thesis of the repre-sentational essence of neo-Brentanian intentionality, this moderate perspective

42 See note 26.

114 JEAN-MICHEL ROY

offers a different way out of the dilemma, which consists in questioning thelimits of these implications without questioning their reality. This perspectivealso offers, thereby, the stimulating hope of overcoming what looks like an irre-ductible opposition. Accordingly, it deserves particular attention.

There are two possible ways of conceiving this compatibilist solution. It canbe firstly understood as locating the error of neo-Brentanian intentionalismnot in its representationalist analysis of objectivation, but in its very assimila-tion of intentionality with such a property. Accordingly, a non-neo-Brentanianapproach to the property of intentionality will be one that refuses such anassimilation and as a consequence severs its ties with the property of represen-tation. It is thus entirely based on the assumption that it is indeed possible toeliminate from the notion of intentionality that of relating to something as anobject, and to do so in a way that also eliminates all connections with the ideaof specification as previously explained. The first aspect of this double assump-tion seems, however, to be particularly challenging and stretches theoreticalimagination beyond limits. The idea of objectivation appears to be so deeplyingrained in that of intentionality that one can indeed reasonably wonder:What could intentionality be if not, one way or the other, objectivation? In theend, terminology is free, and it is certainly possible to extend the term to any-thing we want. But the intentionalism resulting from such an extension wouldobviously run the risk of being entirely artificial and, appearances notwith-standing, no different in fact from intentional eliminativism. And no bettereither. For it is a real question to determine how far a theory of cognition cango without calling upon the idea of objectivation.

According to the second interpretation, the search for a non-Brentanianconception of intentionality should not dispute its assimilation with objectiva-tion, but its understanding of objectivation itself, seen as the root of its com-mitment to representationalism. The key to escaping the dilemma is to find away of elaborating a non-representational analysis of objectivation. This isan undoubtedly more promising approach, although a quite challenging onefrom a theoretical point of view. To what extent, indeed, is it really possibleto dissociate the notion of objectivation from that of representation under-stood in its full extent and radicality, that is to say, as specification? Such iscertainly the most central and deepest issue raised by any attempt to devise anon-neo-Brentanian kind of intentionalism.

And here again there are two ways of understanding the issue and,therefore, of searching for a solution to it. One possibility is that someerror affects the neo-Brentanian analysis of objectivation and that this errorhas as a consequence the additional and different error of giving in to repre-sentationalism. The difficulty in this perspective is to identify this error andto correct it in a way that blocks its representationalist consequences. But

115THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

representationalism remains somehow external to the neo-Brentanian con-ception of objectivation. The alternative is to see it, on the contrary, as builtin this conception itself, making neo-Brentanian objectivation an intrinsicallyrepresentational property. This is what the thesis of the representationalessence of neo-Brentanian intentionality claims. What it asserts is that neo-Brentanian intentionality is revealed to be essentially representationalbecause it assimilates intentionality with objectivation, and because, in turn,objectivation is revealed to be a particular form of representation understoodas specification. According to neo-Brentanism, a mental state is intentionalnot only insofar as it relates to something as an object, but to the extent alsothat it specifies, and hence characterizes or interprets, something as such—afeature that can be located in the elusive component of directedness of Bren-tano’s original definition as the idea that being directed toward something asan object seems to introduce into it a sort of attitudinal element. In this per-spective, the difficulty is to determine to what extent and how the idea ofrelating to something as an object can be dissociated from that of specifyingsomething as an object. In other words, to what extent and how the idea ofobjectivation can be dissociated from that of directedness, at least in thesense of directedness that can be attributed to Brentano.

When understood in this way, searching for a non-neo-Brentanian prop-erty of intentionality as an instrument for making anti-representationalismcompatible with intentionalism looks like a less exotic undertaking than itseemed to be at first glance. But the depth of the requirement should not bemissed. To meet it, it is not enough to simply go for an embodied, extended,enactive . . . form of specification. For such options remain within the pur-view of a representationalist model of the mind correctly apprehended. Simi-larly, it is not enough either to simply devise a notion of intentionality thatbreaks away with the idea of specification. What is necessary is to devise onethat keeps the full theoretical relevance of the property of intentionality for atheory of the mind. As a matter of fact, it is fairly probable that the idea ofrelating to something as an object can be captured to a certain extent inbehavioral terms, as corresponding, for instance, to a specific pattern ofbehavior. But how much of an intentional model of the mind can be devel-oped with such a definition of intentionality? Here is the real question.

7. FORAYS INTO A CRITICAL EXAMINATION OFALTERNATIVE MODELS OF THE MIND

7.1. Defining a critical strategy

A safe and fruitful strategy, in order to come to terms with the specific issuethat has now emerged as the heart of the theoretical dimension of the

116 JEAN-MICHEL ROY

problem of anti-representationalist neo-Brentanian intentionalism, is to startby examining whether the on-going search for alternative models of themind already contains explicit or implicit solutions to it and how acceptablethese are. This strategy consists in nothing else, in fact, than turning to ques-tions Q1 and Q2 which characterize the critical dimension of this problem(see above in Section 1).

It is important to note that the thesis of the representational essence ofneo-Brentanian intentionality comes with a prediction regarding the answersthat these questions can in principle receive. This prediction is that if one ofthe alternative models of the mind under development advocates a radicallyanti-representationalist form of neo-Brentanian intentionalism, then it mustfall into one of the following theoretical categories:

1. It really advocates a neo-Brentanian and radically anti-representationalist kind of intentionalism, but is contradiction riddenand then not acceptable;

2. It really advocates a neo-Brentanian kind of intentionalism, but provesto be in fact not radically anti-representationalist, and is thenacceptable;

3. It really advocates a radically anti-representationalist kind of model ofthe mind, but proves to be eliminativist about intentionalism, and isthen acceptable;

4. It really advocates a radically anti-representationalist kind of intention-alism, but of a non-neo-Brentanian and, hence, not essentially repre-sentational type, and is then acceptable.

In terms of this prediction, the strategy can be described as one of adopting alimited approach to the critical dimension of the problem that focuses on theidentification of possible representatives of the fourth one of these categoriesamong protagonists of the current search for alternative models of the mind.

In spite of offering a possible shortcut for solving the core theoretical issue,this critical strategy raises a methodological difficulty: where to look for suchrepresentatives in priority? A reasonable answer is to privilege alternativemodels of the mind with explicit claims pointing, in varying degrees, in thedirection of an anti-representationalist form of intentionalism. These modelscan, in turn, be distributed into four descriptive categories of decreasing pri-ority. First are the models that explicitly advocate a non-neo-Brentanianform of intentionalism and might even ground their anti-Brentanism in anti-representationalism. Second are the models that advocate a form of anti-representationalist intentionalism without being explicitly directed against

117THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

neo-Brentanian intentionalism. A third convenient category includes thosethat defend intentionalism at large without adopting any particularly mani-fest position about representationalism. And fourth are those that, on thecontrary, defend anti-representationalism at large without adopting any par-ticular manifest position about intentionalism.

A full application of this research strategy is obviously beyond the scope ofthe present study. But its fruitfulness can be illustrated with a preliminaryinvestigation of two particularly revealing cases.

7.2. Rowland’s Neo-Neo-Intentionalism

Among present day alternative models of the mind, the “new science of themind” proposed by M. Rowlands is probably unique in the emphasis it putson the issue of intentionalism, which it considers as the most basic one aboutthe foundations of cognitive science. Indeed, in the 2010 synthesis of hisviews published under this title, Rowlands pleads in a rather standard wayfor a form of anti-Cartesian explanation of cognition, labelled “the amalga-mated mind thesis” and that combines a version of vehicle externalism with aversion of cognitive embodiment. But his most crucial claim is that this amal-gamated mind thesis is entirely rooted in the intentional character of cogni-tion. In others words, his key point is that the explanation of cognition mustgo anti-Cartesian because it must be intentional. In Rowlands’opinion, theerror of the science of cognition born with the cognitive revolution is not thatit resuscitated intentionalism, but that it gave to its neo-intentionalism anunduly neo-Cartesian orientation, which must be discarded on the basis ofan “alternative conception of intentionality.” An alternative conception thathas “the amalgamated mind . . . as an obvious . . . consequence”43 and is notunheard of, but has not been sufficiently taken into consideration in the past.

Rowlands provides, accordingly, a perfect example of an attempt to pre-serve intentionalism within the search for an alternative cognitive science of anon-Cartesian kind. To this end, moreover, he explicitly intends to substitutethe neo-intentionalism that prevailed up to and through the neurocognitiveturn with what might be termed a “neo-neo-intentionalism.” This renewedintentionalism offers, therefore, a serious chance of being one that might breakaway from the representationalist characterization of intentionality as anobjective form of specification. All the more so since Rowlands also develops,like most other partisans of cognitive anti-Cartesianism, a strong critique ofthe relevance of the property of representation to cognitive explanation. Apressing question is therefore raised: Does Rowlands’ alternative conception of

43 Rowlands 2010, p. 189.

118 JEAN-MICHEL ROY

the nature of intentionality manage to break away from such a representation-alist characterization? And if so, how and with what success?

The answer is negative and without much possibility for appeal, for hisalternative conception turns out to be even more committed to the idea ofobjective specification than the neo-Brentanian one. Rowlands’ neo-neo-intentionalism is in reality a neo-neo-Brentanian intentionalism; it is a neo-Brentanian one whose Brentanian nature is thrown into full light.

The preliminary description of the phenomenon of intentionality that ittakes as a starting point is rather telling in this regard: intentionality is pri-marily assimilated with a “directedness toward objects” in the purest Brenta-nian tradition, with the notion of directedness occasionally paraphrased interms of reference and aboutness. Furthermore, the next step of the inquiryconsists in reviving a “standard model” of its nature, which is no less rootedin that tradition and finds its direct origin in Meinong. It analyzes intention-ality into a mental act, an intentional object, a transcendant one, and a con-tent whose role is to describe and thereby determine the intentional object.

The innovative feature that the alternative conception Rowlands claims tointroduce in this model, which turns out to be a fairly conservative one, con-sists in giving an additional function to the content beyond its descriptiveone, namely the “transcendantal role”44 of making directedness towardsobjects possible. For a content, in Rowlands’ eyes, is something that not onlyspecifies what the intentional object is, but more fundamentally makes itemerge as such, and this in a “constitutive rather than a causal sense.”45

Which is to say, the intentional object is established as intentional objectthrough the description of it that the content provides. But this idea is in factnothing else than a declination of the idea that intentionality is at its coreobjectivation. As a matter of fact, Rowlands writes: “As transcendental, expe-riences [i.e., contents in this context] are not objects of awareness but that invirtue of which objects . . . are revealed to a subject precisely as objects of thatsubject’s experience.”46 The innovative feature of his alternative conceptionis therefore only one additional step into a fully explicit Brentanian concep-tion of intentionality as objectivation, even though Rowlands prefers to cap-ture it in terms of disclosure: “The essence of intentionality,” he also writes,“is disclosing activity.”47 And, this disclosing activity is not only one by virtueof which the subject relates to the world as an object, but one also thatbestows on it this objective status, which treats or interprets it as such.

44 Rowlands 2010, p. 185.45 Rowlands 2010, p.194.46 Rowlands 2010, p.169.47 Rowlands 2010, p. 195.

119THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

Rowland’s identification of intentionality with disclosure is not only anexplicitation of the long running assimilation of intentionality with objectiva-tion, but also with objectivation understood as objective specification.

His real departure from this traditional and prevailing conception is to befound, rather, in the modalities through which he sees this disclosing activity tobe exerted. Athough he remains faithful to the naturalist reinterpretation ofBrentano’s thesis, Rowlands thinks that the implementation basis of this objec-tive specification involves not only the neural processes of the brain, but alsothe activity of the body and the manipulation of external structures. Further-more, as already indicated, his key point in this respect is that when its natureas objective specification is thrown into full light, such an extension of thescope of the natural realizers of intentionality is revealed to be a necessity.

By subscribing to a form of intentionalism that turns out to be an amalga-mated version of a neo-Brentanian one well understood, Rowlands is never-theless also fundamentally subscribing to representationalism. Should histheory be put, as a result, into the category of contradiction-ridden alterna-tive models of the mind? Or instead into the category of models with merelythe appearance of radical anti-representationalism? Neither of them, in fact,because the anti-representationalism he advocates is a limited one and is soon three different counts. It is firstly limited in the sense that it does not rejectentirely the relevance of the property of representation, but only argues forits dispensability in many cases where it was thought to be necessary. It islimited secondly in the sense of being mainly directed against internal repre-sentations and not external ones. Finally, it is also limited in the sense ofbeing based on a stand in conception of representation, even though itrightly emphasizes the specification dimension of representations by charac-terizing them primarily as entities that say something about the world. As aresult of these limitations, and of the third one in particular, Rowlands canlegitimately make room in his theory, at some level of analysis, for the ideathat some mental states are intentional without being representational. Theopposition between representational and non-representational intentionalityis essentially for him one between a disclosure performed by means of aninformation-bearing structure and one not performed by means of such astructure. But, by virtue of the understanding of intentionality as objectivespecification that he implicitly shares, this non-representational intentionalityremains, at a deeper level of analysis, representational in a different andmore basic sense of the term, namely that of objective specification. And,without contradiction, stand in representation is a species of objective specifi-cation. However, Rowlands’ neo-neo intentionalism fails thereby to show ushow the notion of objectivation can be appropriately dissociated from that ofspecification. If anything, he reasserts the strength of their association.

120 JEAN-MICHEL ROY

7.4. The Anti-Representationalism of Radical Enactivism

Among contemporary alternative models of the mind, the closest thing to anauthentically non-representationalist intentionalism is not to be found inneo-neo-intentionalist ones of a different vein, as most of them seeminglyshare the same limitations as Rowlands’,48 but, rather, in certain forms ofradical anti-representationalism associated with enactivism, in particularwith the kind of enactive approach defended by Daniel Hutto and Eric Myinin their already quoted Radicalizing Enactivism.

As clearly suggested by the very title of the book, these authors see theirtheory of the foundations of cognitive science as belonging to the enactiveperspective opened by the Embodied Mind of Franciso Varela, Eleanor Rosh,and Evan Thompson,49 but as offering a more radical version of it than theone with which it was inaugurated, that they label “autopoietic enactivism,”and as more radical as well than the alternative formulation it was providedby Alva No€e and Kevin O’Regan under the name of sensorimotor theory. Intheir eyes, enactivism at large is essentially an anti-representationalist view ofcognitive explanation, whose main claim is that the most basic aspects of cog-nition must be accounted for in terms of an activity not of forging andmanipulating representations located in the brain, but of an interaction ofthe whole organism with its environment to be analyzed with the sole con-ceptual tools of dynamic systems theory. Such a claim that comes down, intheir opinion, to rejecting the classical assumption that the property of (rep-resentational) content belongs to the essence of cognition, which they see asdefining what they dub intellectualism in the area of cognition.50 Further,what makes radical enactivism radical in their eyes is nothing more than therigor with which it deploys this fundamental and double-sided claim,51 recti-fying the deviant compromissions with intellectualism that the other versions,including the founding one, indulge in.52

Although limited to basic cognition, this uncompromising elimination ofthe property of representation is all the more relevant to the present inquirysince it rests on an understanding of that property that, as previously under-lined, is in full accordance with the one articulated here. Indeed, Hutto and

48 Although, unfortunately, I cannot do justice to it, this criticism applies in my opinion tothe theory of operative intentionality laid out by Shaun Gallagher and Katsunori Miyahara,which constitutes another particularly significant development of a neo-neo-intentionalism inthe context of the search for alternative models of the mind, as it is intended as an explicitattempt to devise a concept of intentionality that is consistent with the alternative concept ofmind suggested by the enactive and extended approaches (see Gallagher et al. 2012).

49 See Varela et al. 1993.50 Hutto 2013, p. 9.51 Hutto 2013, p. 11.52 Hutto 2013, p. 24, 34.

121THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

Myin formulate their rejection of representationalism in simple and basicterms that fully recognize the fact of attributing determinations or specifyingas the most crucial feature of representing, even though they also offer adetailed criticism of the various contemporary reconstructions of representa-tionalism based on the notion of informational content. They write, forinstance, that the nonrepresentational character of organismic activity lies inthe fact that this activity “does not depend upon individuals retrieving ofinformational content from the world . . . in order to attribute properties tothe world.”53 An essential characteristic that what they call “maximally mini-mal representationalism” captures most adequately when it claims with TylerBurge that “perceptual experience represents if and only if it takes some por-tion of the world to be a certain way [,] if it attributes properties to theworld.”54 And, that is nevertheless not disconnected from the standard oneof possessing conditions of satisfaction, still considered by radical enactivismas “the most general and minimal requirement on the existence of con-tent,”55 and seemingly implied by it. An implication that can take differentforms, in particular that of possessing truth conditions or, more simply, accu-racy conditions.56

Even more interesting, however, is the fact that this anti-representationalism, primarily and correctly put in terms of rejection of theproperty of specification, is complemented with an explicit acceptance of therelevance of the property of intentionality in the explanation of basic cognitionitself. Thoughout their book, Hutto and Myin assert again and again that,despite being “nonrepresentational and content free,”57 organismic activity is“intentionally directed,” that organisms exhibit “determinate kinds of inten-tional directedness towards aspects of the environment,” and that these “basicforms of directedness” do not require to be analyzed in “semantic, contentfulor representational terms.”58 In other words, their “radical enactivism”embodies an unambiguous pretention to offer an intentionalism based on anauthentically nonrepresentational understanding of the property of intention-ality. The whole question is therefore to clarify what this nonrepresentational-ist property of intentionality really amounts to, and to what extent inparticular it preserves an essential link with the notion of objectivation. But,the truth is that the situation is far from being clear in this respect.

53 Hutto 2013, p. 5.54 Hutto 2013, p. 102.55 Hutto 2013, p. 102.56 Hutto 2013, p. 103.57 Hutto 2013, p.13.58 Hutto 2013, p.78.

122 JEAN-MICHEL ROY

It should first be emphasized that it seems safe from re-introducingthrough the back door the characteristics of specification that it officiallyrejects by being essentially characterized as a certain form of behavioralresponse. “Experiencing organisms,” Hutto and Myin write, “respond to[certain worldly] offerings in distinctive sensorimotor ways that exhibit acertain minimal kind of directedness.”59 But in what does this directednessof behavior consist, if it cannot be in “inherently say[ing] anything aboutthat how things stand in the world”? If it corresponds to a nonrepresenta-tionalist account of the property of objectivation, it should in fact reduce toa form of behavior that consists in responding to “certain offerings of theworld” as objects, in contrast to responding to certain offerings of the worldas simple things. That the notion of directedness should mean objectivationfor Hutto and Myin is supported by the fact that they claim, in addition,that this notion is derived from the doctrine of teleofunctionalism. It is, soto speak, what is left of teleofunctionalism when it is dissociated from itsrepresentational component. Although it is not so clear what such a sub-straction of representationalism from teleofunctionalism really means, itclearly points in the right direction, since, as a theory of representation, tel-eofunctionalism is a theory of objectivation. But if this is indeed what thenotion of intentional directedness means for Hutto and Myin, the concretecharacteristic of such respondings must be specified. In what does respond-ing behaviorally to something as an object correspond exactly behaviorallyspeaking? The theory remains silent on this point. Furthermore, it is noteven clear that the directedness of behavior should be interpreted in thisway, as Hutto and Myin remain almost perfectly silent as well on how itshould be understood. And, some of their phrasings even point in the direc-tion of another interpretation, according to which it corresponds to nothingmore than the fact, for organisms, of being primarily and by virtue of theirown nature responsive to certain elements of their environment. In thiscase, their nonrepresentationalist property of intentionality reduces to a sortof property of natural attunement and thus loses its connection with that ofobjectivation, suggesting, thereby, skepticism as to the possibility of obtain-ing a property of objectivation of a nonrepresentational kind.

REFERENCES

Bechtel, W. 1998. Representations and cognitive explanations: assessing the dynamicistchallenge in cognitive science. Cognitive Science 22.

59 Hutto 2013, p.19.

123THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM

Betti, A. 2013. Judgement and truth in early analytic philosophy and phenomenology.In Judgement and truth in early analytic philosophy and phenomenology, ed. M. Textor. Lon-don: Palgrave.

Brentano, F. C. 1874/2009. Psychology from an empirical standpoint. New York: Routledge& Kegan Paul.

Brentano, F. 1930/1966. The true and the evident. New York: Routledge & Kegan Paul.Brooks, R. 1991. Intelligence without representation. Artificial Intelligence 47: 67–90.Chisholm, R. M. 1952. Intentionality and the theory of signs. Philosophical Studies 3: 56–62.Fodor, J. A. 1987. Mental representation: An introduction. In Scientific inquiry in philo-

sophical perspective, ed. N. Rescher. Lanham, MD: University Press of America.———. 1987b. Psyhosemantics. Cambridge, MA: MIT Press.Goodman, N. 1976. Languages of art: An approach to a theory of symbols. Indianapolis, IN:

Hackett.Gallagher, S. and K. Miyahara (2012). Neo-pragmatism and enactive intentionality. In

Action, perception and the brain: New directions in philosophy and cognitive science, ed. J.Schulkin. Houndmills, Basingstoke, Hampshire, UK: Palgrave Macmillan.

Grush, R. 1997/2001. The architecture of representation. In Philosophy and the neuroscien-

ces: A reader, eds. P. M. William Bechtel, Jennifer Mundale, 5–25. Oxford: BasilBlackwell.

Haugeland, J. 1990. The intentionality all-stars. Philosophical Perspectives (Action Theory and

Philosophy of Mind) 4: 383–427.———. 1998. Having thought: Essays in the metaphysics of mind, ed. J. Haugeland. Cam-

bridge, MA: Harvard University Press.H€ofler, A. and A. Meinong. 1890. Logik. Journal Philosophische Propadeutik 1.Husserl, E. 1900–1901/2000. Logical investigations. Amherst, NY: Humanity Books.Hutto, D. D. and E. Myin. 2013. Radicalizing enactivism. Cambridge, MA: MIT Press.No€e, A. 2004. Action in perception. Cambridge, MA: MIT Press.Rowlands, M. 2010. The new science of the mind: From extended mind to embodied phenomenology.

Cambridge, MA: MIT Press.Roy, J.-M. 2001. L’�emergence de la neuroscience cognitive. Cahiers Alfred Binet 667.———. 2011. From cognitive Cartesianism to cognitive anti-Cartesianism: A hypothe-

sis about the development of Cognitive Science. In New Perspectives on the history of

cognitive science, eds. P. Csaba, L. Gurova, & L. Ropolyi. Budapest: Akad�emiaiKiad�o.

Russell, B. 1913/1984. Theory of knoweldge: The 1913 manuscript. London: Routledge,1992.

———. 1918. Mysticism and logic. London: Unwin Books, 1974.Searle, J. R. 1983. Intentionality, an essay in the philosophy of mind. Cambridge Cambridge-

shire; New York: Cambridge University Press.Spiegelberg, H. 1970/1981. “Intention” and “Intentionality” in the scholastics, Bren-

tano and Husserl. In The context of the phenomenological movement, ed. H. Spiegelbreg,189–216. The Hague: Martinus Nijoff.

Textor, M. Judgement and truth in early analytic philosophy and phenomenology.Twardowski, K. 1894/1977. On the content and object of presentations: A psychological investiga-

tion. The Hague: Nijhoff.

124 JEAN-MICHEL ROY

van Gelder, T. 1995. What might cognition be, if not computation? The Journal of Phi-losophy 91: 345–81.

Varela, F., E. Thompson, et al. 1993. The embodied mind. Cambridge, MA: MIT Press.Wheeler, M. 2005. Reconstructing the cognitive world: The next step. Cambridge, MA: MIT

Press.

125THE PROBLEM OF ANTI-REPRESENTATIONALIST INTENTIONALISM