Being a Direct Realist – Searle, McDowell, and Travis on ‘seeing things as they are’

Abstract

The aim of the present article is to identify and analyze three particular disputes among current proponents of perceptual realism which may throw light on tensions present in the history of direct realism and current discussions. Starting from John Searle’s conception of direct realism, I first set McDowell and Travis’s approaches in contrast with it. I then further compare Travis’ view with McDowell’s. I claim that differences among the three philosophers are traceable first to methodological conceptions of the approach to perceptual experience (whether philosophical naturalism implies dealing with the sub-personal level), then to what makes for the particularity of a perceptual experience (whether it involves consciousness and a task of unity or not), and finally to what makes for the determinacy of an experience of things in the world (whether such determinacy characterizes the world itself or, as such, involves language and thought).

1 Introduction

Borrowing John Searle’s words, direct realism about perception is about ‘seeing things as they are’, without any mediation of concepts, ideas or representations. The direct realist should not be committed to any kind of interface between the perceiver and whatever is perceived¹. My leading question in this article will be what, exactly, one is claiming about perceptual experience and about the world in saying that. After all, every direct realist intends to give us a philosophical picture of what seeing things as they are really is, and, still, very different pictures come out of such effort. In this article I will go through the positions of three philosophers who regard themselves as realists, arguably direct realists, regarding perception: John Searle, John McDowell, and Charles Travis (Searle 2015, McDowell 2013a, b, Travis 2013, 2013/2004). I intend to show that the very deep divergences between their views of perception start very early in the game: in fact, they start with what each one takes the actual philosophical issues concerning perception to be, and what issues regarding perception each one takes to be properly philosophical. One main purpose in what follows will thus be to bring forth what Searle, McDowell, and Travis think ‘seeing things as they are’ really is. That will then make it possible to identify where they part ways in their direct realist approaches to perception. Hopefully, this will not only illuminate the history of direct realism but also bring out challenges for the present.

2 Searle on ‘seeing things as they are’: seeing-as and seeing-that

John Searle’s 2015Seeing things as they are is a book-length defence of direct realism². It is aimed not so much against classic sense data theorists but mostly against (supposed) current opponents of direct realism such as ‘the disjunctivists’, by which he means especially John Campbell and Mike Martin, ‘the weeds growing in his own garden in Berkeley’ (see Searle 2015: 8, and Searle 2015:166). One of Searle’s main contentions in the book is that the sense data theorists and the disjunctivists share allegiance to a ‘Bad Argument’ (I will consider this shortly). So what are, according to Searle, ‘things as they are’, and what is seeing them thus? Searle takes a revealing step in Chapter 2 of the book (The intentionality of perceptual experiences’, Searle 2015:74) when he briefly brings in the duck-rabbit, a well-known ambiguous image Wittgenstein uses in the Philosophical Investigations (Wittgenstein 2009, Part II, § 118), in his discussion of seeing aspects. In Searle’s view, what is remarkable about the duck-rabbit is that one can have different experiences produced by the exact same stimulus without any phenomenon of illusion, hallucination or delusion occurring. One sees the duck-rabbit as a duck, and then one sees the duck-rabbit as a rabbit. According to Searle, this happens because of the intentionality of visual experience. It would be natural to think that the duck-rabbit is a very special case of seeing-as. That is not what Searle thinks, though. What he ends up defending in Seeing Things as They Are is that all seeing is seeing-as, exactly as is the case with the duck-rabbit. In his own words, «All visual intentionality is a matter of presentations, presentations are a subspecies of representation, and representation is always under some aspect or other. » (Searle 2015: 75). In other words, for Searle, all visual experience, and not just apparently special cases like the duck-rabbit, is a matter of seeing-as. The most common, ordinary, visual experiences are a matter of seeing-as. My seeing you now, if I were to face you, or my looking out of the window and seeing cars or buildings, are cases of seeing-as for Searle. So seeing-as is not a phenomenon restricted to specific cases such as the duck-rabbit or the other cases Wittgenstein uses in the Philosophical Investigations, such as seeing a similarity between two faces (Wittgenstein 2009, Part II, § 113), or seeing a triangle as a triangular hole, or as a solid, or as standing on its base, or as hanging from its apex, or as an arrow or a pointer, or as pointing downwards or upwards (Wittgenstein 2009, Part II, §162), or seeing a figure as a glass cube or as an open box or a wire frame (Wittgenstein 2009, Part II, § 116). Such examples are not what Searle has in mind. For him, there is no interesting contrast between them and cases of looking out of the window and seeing cars or buildings, or ducks, or rabbits. He insists that all seeing is seeing-as, or, we might say, all seeing is taking a this to be thus³. For Wittgenstein himself this makes no sense. It makes little or no sense to say at the sight of a knife and fork ‘Now I see this as a knife and fork’. As he puts it, «This utterance would not be understood. Any more than “Now it is a fork for me” or “It can be a fork too”. » (Wittgenstein 2009 § 122). In Wittgenstein’s mind such cases of seeing are precisely to be contrasted with the duck-rabbit. In the case of the duck-rabbit I, as it were, exclaim to myself “Now I see it as a rabbit” or “Now I see it as a duck” and that is clearly not the case when I face a real-life duck, or a real-life rabbit, or a knife, or a fork. Yet for Searle all cases of real-life common seeing are cases of seeing something under an aspect. This is the core claim of what I want to call Searle’s representationalism. His long-held view of intentionality is marked by such representationalism (see Searle 1983, Chapter 2 Intentionality and Searle 2015, Chapter 2 The Intentionality of Perceptual Experiences). According to Searle, intentionality is that by which the mind is directed at objects and states of affairs in the world (or is about, or of, them). Like Edmund Husserl in the early 20^th century, when setting the stage for philosophy done as phenomenology, Searle explores a view of the mental in terms of intentionality (Husserl 2001). There is one all-important difference though. In plain contrast with Husserl, Searle sees intentionality as a biological phenomenon, occurring inside a living organism, and thus common to humans and other animals. As such, «the most basic forms of intentionality are the biologically primitive forms such as conscious perception, intentional action, hunger, thirst, and such emotions as anger, lust and fear» (Searle 2015:33). In other words, Searle’s view of intentionality is openly naturalistic: this is why the contrast with the beginnings of phenomenology in the hands of Husserl is relevant. Searle’s view is naturalistic precisely in the sense that Husserl aimed to criticize and reject. Husserl believed naturalism led to psychologism, which he saw as the wrong path for philosophy (Miguens 2010). Phenomenology was devised by Husserl as a non-psychologistic approach to the mental. For Husserl if an event occurring, say, in the visual cortex is not part of the flow of consciousness (in the sense that a particular biochemical or electric phenomenon occurring in the brain is not itself part of the flow of consciousness), then it is not an object for philosophy practiced as phenomenology. Philosophy practiced as phenomenology, as intended by Husserl, aims at describing consciousness only, or more specifically, as Husserl put it, essences in consciousness. In other words, phenomenology is an analysis of meaning, of objects of thought and experience (Miguens 2010, 2018, 2019)⁴. Phenomenology does not take as its object occurrences in the natural world such as events in the brain. That is the job of science. Yet Searle’s approach to the mind includes the physical taking place of the phenomena he regards as intentional. He always insists that intentional states are caused by the brain and realized in the brain, and that this is of interest to the philosopher.

I evoked the historical beginnings of phenomenology in Husserl’s hands to identify one first decision concerning a philosophical approach to perception and what being a direct realist about perception involves. For a direct realist such as Searle, a philosophical approach to perception concerns not only phenomenology (which he takes to mean ‘things seeming to me a certain way’ and thus something quite different than what ‘phenomenology’ means for Husserl) but also the sub-personal vehicles of cognitive phenomena. This position then sets the tasks for the philosophy of perception: philosophy of perception deals with issues of causality and physical realization concerning such sub-personal vehicles in the same breath as questions concerning seemings. Such a naturalist approach to perceptual experience (and what is thereby assumed regarding philosophical method) is clearly at odds with Husserl’s conception of phenomenology. Now the two realists concerning perception I consider next, McDowell and Travis, would squarely side with Husserl in thinking that the concern of a philosophical view of perception, unlike the concern of perceptual science, is not the sub-personal goings-on within a cognitive system. The concern of a philosophical view of perception is rather perceptual experience and its relation to the world and to phenomena of meaning and knowledge. Nothing in this per se speaks against the brain being part of the world and there being a natural science approach to it, of course. Yet such difference is crucial for putting Searle’s next step in perspective. In the framework of his sort of naturalism, Searle works with the idea of conditions of satisfaction. The idea of conditions of satisfaction is key to understanding intentionality in general. There he seems to be pursuing the path of semantic analysis of Husserl, or Frege. Granted, he carefully distinguishes perceptual experiences and beliefs in this respect. He sees beliefs as the most philosophically interesting of intentional states (Searle 2015: 34) because they have whole propositions as contents. This makes for a contrast between beliefs and experiences: «Unlike [what is the case with] our beliefs and statements, we do not say that our visual experiences are true or false» (Searle 2015: 57). I will get back to that. Searle does think that beliefs are philosophically interesting because they have entire propositions as contents, and he is careful to point out that beliefs are not attitudes towards a proposition. Here we get to the most essential point when we are thinking about intentionality according to Searle. What is essential here is that we should distinguish between content and object – this is where his critique of those philosophers of perception who are not direct realists of his sort rests. Beliefs (or rather, most beliefs⁵) are not about propositions. Propositions are the contents of beliefs or a way to speak of the content of beliefs, not the objects of beliefs. The same content-object distinction applies to perceptual experience. If I see a computer in front of me, the content of my perceptual experience is that there is a computer in front of me. The object is the computer itself. «One never just sees an object», Searle, the direct realist, claims (Searle 2015: 110). Admittedly, a surprising claim for a direct realist. The way Searle puts things, what I see is that there is a computer in front of me, not the object. That one never just sees an object is what allows Searle to claim that the content may be the same in a case of veridical perception as in a corresponding hallucination – the content may be the same, but not the object. This is how Searle gets to the Bad Argument, the main target of his 2015 book. The Bad Argument is the systematic confusion of content and object when thinking about perception. It amounts to mistaking the content of a perceptual experience for its object. This, according to Searle, is a faux pas common to many people, from sense data theorists to disjunctivists such as Mike Martin, or John McDowell. The best prevention against the mistake, according to Searle, is to stick to the idea that one doesn’t experience perceptual contents, one has them.

As for perceptual experiences, they contrast with beliefs since theirs is a specific form of intentionality. They have a different phenomenology (in Searle’s sense of ‘things seeming to me a certain way’). Perceptual experiences have phenomenal properties. They are not just representations but also presentations. When Searle describes visual phenomenology, he says that it extends from the top of my forehead down to my chin. He calls it an ‘area’: he says that I don’t have visual consciousness behind my head or under my feet (Searle 2015: Introduction). There is a locus of visual consciousness. I have it in front of my head even if my eyes are closed (and this is phenomenology, not physiology, he stresses). This area is what he calls subjective visual field. So, for Searle visual experience (‘seeing things as they are’) is a subjective experience that takes place inside the head. If I open my eyes the subjective visual field fills up, since I become visually aware of the objective visual field. The objective visual field he identifies with objects and states of affairs around me. Searle insists much on this difference between subjective visual field in your head and objective visual field outside your head. His 2015 account of perception proceeds by telling us how subjective and objective visual fields relate. The key idea for this relating is that all seeing is intentional: «The fact that the processes in the subjective visual field, the experiences, have intentionality has two important consequences (…): all seeing is seeing-as and all seeing is seeing-that.» (Searle 2015: 110).

So, in brief, Searle’s perceptual realism belongs within a naturalistic framework for thinking of the mind qua intentionality. Intentionality, as Searle sees it, characterizes the physical world itself. Physical phenomena may be said to be intentional: it is a characteristic of brain states, namely, that they have intentionality. Intentionality is no mere ‘mark of the mental’ as the influential Brentanian idea for distinguishing mental from physical phenomena had it. Given this, perceiving as the experience of, say, seeing, is seen as taking place inside the head⁶. Because all perceptual experiences are intentional, all seeing is seeing-as, and that is how one gets to ‘outside the head’. Such seeing-as Searle then simply equates with seeing-that. Seeing things as they are is, then, for Searle, seeing-that.

3 McDowell on ‘seeing things as they are’: conceptual capacities, intuitional content and what a ‘sane naturalism’ must be

I will now contrast Searle’s picture of perceptual experience with McDowell’s (McDowell 1994, 2009a, b, c, d, 2013a, b). What is, according to McDowell, ‘seeing things as they are’? The first thing to be clear about is that although McDowell, like Searle, claims to be a naturalist, his naturalism is quite different from Searle’s. According to McDowell, no sane naturalism should be able to persuade us that being in the world is being inside our heads. Perceiving, for McDowell, is simply being in the world (McDowell sometimes traces this idea to Hilary Putnam’s dicta on mind and meaning, that ‘the mind is not in the head’, or that ‘meaning is not in the head’). It is important to be clear about this starting point if McDowell’s motivations for his perceptual and epistemological disjunctivisms are to be properly understood. For McDowell, the gist of disjunctivism is to acknowledge presentness and acquaintance as marks of perceptual experience. The difficulty in doing this has to do with the need to do it while at the same time de-subjectivizing the appearance-reality distinction and taking perception to be a form of being in the world and not of being inside our heads. Such backstage ideas, anyway, explain why McDowell strives to make room for the bearing of the ‘good cases’ in his approaches to perception and to knowledge. He wants to bring forth an asymmetry between such good cases and the ‘bad cases’ such as illusion and hallucination (in the case of perception). I will take a brief look at some important aspects of this picture.

McDowell’s article Criteria, Defeasibility and Knowledge (McDowell 1998) is the right place to start. The article is the locus of the initial rejection of the so-called highest common factor views. The idea of the highest common factor is that there is a highest common factor to good and bad cases, and thus, in the case of visual perception, to veridical perception, illusion, and hallucination. The article is thus a landmark in the history of disjunctivism, the very same disjunctivism which is Searle’s target in his 2015 book. That disjunctivists reject common factor views is one point about disjunctivism that Searle does get right. In the article McDowell looks into the connection between highest common factor views and a particular conception of appearances, which he believes is a bad, ungrounded, conception of appearances, belonging in a particular ‘picture’ of perception (in the Wittgensteinian sense of something which holds us captive in a way of thinking – in this case a bad one). McDowell’s own alternative picture of appearances involves a non-Cartesian, or modest, conception of what ‘indistinguishability’ for a subject is. Such alternative conception is a crucial element of his view of experience. The idea is that my being unable to distinguish a veridical perception from an illusion or hallucination does not allow for any heavy conclusion – it simply marks the limits of my self-awareness (in Mike Martin’s terms). It does not by itself reveal anything like a common factor between veridical perception, illusion, and hallucination (the good and the bad cases).

In Criteria, Defeasibility and Knowledge McDowell is interested in exploring the consequences for epistemology of the ideas above. He focuses on the cases of perception and of knowledge of other minds. In both cases, perception and knowledge of other minds, the bad cases (which are illusion or hallucination and pretense, respectively) are taken to be mere appearances. McDowell thinks there is something wrong in the very setup of the problems here. He believes there is an unacknowledged asymmetry, which he wants to bring out. As he sees things, the good case and the bad case are ontologically and thus epistemologically very different. This is the reason why he suggests one should think of them in terms of a disjunction: «But suppose we say – not at all unnaturally – that an appearance that such and such is the case can either be a mere appearance or the fact that such and such is the case making itself perceptually manifest to someone. As before, the object of experience in the deceptive case is a mere appearance. But we are not to accept that in the non-deceptive case too the object of experience is a mere appearance and hence something that falls short of the fact itself. On the contrary, the appearance that is presented to one in those cases is a matter of the fact itself being disclosed to the experiencer. So appearances are no longer conceived as in general intervening between the experiencing subject and the world.» (McDowell 1998: 80).

In the above we have McDowell, the disjunctivist, showing us why the disjunctivist claims to be a direct realist: he is doing away with a (bad) conception of appearances, a conception that depicts appearances as mediators, as something which intervenes between mind and world. Also, in McDowell, perceptual experience is identified not with ‘being inside the brain’ but rather with being in the world. Such experience can simply be of ‘things being as they are’. In perception things themselves, as it were, are, at least in some cases, being made manifest to the experiencer. That is what perceiving is, not some kind of ‘representation of how things are’. It might still be argued, though, that McDowell’s way of doing away with a common factor and claiming that experience does not ‘fall short of the fact itself’ still appeals (differently) to mediation. His view of seemings brings in a sort of representation-like ‘entities’, and so a form of representationalism. Granted, such representationalism is very distant from Searle’s representationalism. Still, it is a very important element in the full picture of McDowell’s perceptual realism.

I will look into the more recent McDowell, in Avoiding the Myth of the Given (McDowell 2009c), as he considers perceivings and seemings to substantiate this claim. In other pieces in his Having the World in View (McDowell 2009a, b, d) and in (McDowell 2013a, b) he explores details and consequences of this position. For McDowell, a perceiving is a taking-to-be, which then is (or is not) endorsed by judgment. Perceivings (seemings) are claim-like, although such claims are not yet judgments. Experiences (the perceivings) are said to make claims in that they reveal (immediately, in intuition) things to be the way they would be judged if they were to be judged. ‘Claim’ is a term of Sellars’, which McDowell thinks is ‘wrong in the letter but right in spirit’ (McDowell 2009c, 267). There is one important motivation here: in articulating his view of perception McDowell is keeping his eyes on epistemology and on questions of justification and warrant (Miguens 2020). An account of perceptual experience is key to an account of knowledge – to what counts as knowing that things are thus and so. Now if an account of experience is to be key in an account of knowledge, it is important to have in mind how any judgments, and so also perceptual judgements, can be warranted. When we speak of judgments we speak, according to McDowell, of conceptual capacities being exercised, and any talk of conceptual capacities belongs in the logical space of reasons. Judgments can be warranted only in terms of relations within the space of reasons (this is how McDowell puts things in his classic Mind and World, McDowell 1994). Only representations enter relations within the space of reasons. If this is so, then taking perceptual experiences to be in a specific sense representational is crucial for an account of knowledge. Granted, McDowell has reformulated his Mind and World position (e.g., in Avoiding the Myth of the Given). When he, in Mind and World, spoke of representational content that amounted to a commitment to propositional content. He accepted the idea that experience has representational content, tout court. In Avoiding the Myth of the Given he does not claim anymore that perceptual seemings have propositional content. He now claims that only judgments and assertions have propositional content. He still claims, though, that perceptual experiences have content: they have what he calls intuitional content. McDowell’s intuitional content is an interpretation of Kant’s Anschauung; it is in fact the term Anschauung that McDowell translates as a ‘having in view’ (McDowell 2009a, b, c, d, e). It is because perceptual experiences have intuitional content that they are representational. In fact, McDowell believes that if we do not acknowledge that perceptual experiences are in this way representational, we succumb to the Myth of the Given. Put in Kantian terms, we succumb to the Myth of the Given if we do not acknowledge that the Understanding is at play in Sensibility itself. Or, we could also say, that we succumb to the Myth of the Given if we believe that we can keep the workings, as it were, of Sensibility and of the Understanding conceptually separate in an account of perception. We cannot, according to McDowell. This is where, then, his direct realism meets Kant⁷. At this point, in appealing to Kant, McDowell does not necessarily care much about Kant’s global transcendental framework, or Kant’s global view of subjectivity (including, say, the difference between understanding and reason, or the Critique of Pure Reason inventory of the specific forms of propositional unity, i.e., the categories listed, see Kant 1998). What he in fact cares about is unity, the unity of judgement as it relates to intuition. Hence the quote from the Critique of Pure Reason he often resorts to: «The same function which gives unity to the various representations in a judgement also gives unity to the mere synthesis of various representations in an intuition; and this unity in its most general expression, we entitle the pure concept of the understanding» (Kant 1998, Critique of Pure Reason A 79/B104-105, § 10 Transcendental Analytics).

In McDowell’s terms, we fall prey to the Myth of the Given by not acknowledging that only the unity of (content) of judgment makes for the unity of content of intuitions. Such is McDowell’s interest. This is why for him the workings of sensibility and of the understanding cannot be separated in an account of perception. Now in Kant himself, at this point, the issue seems to be not only sensibility and the understanding and keeping or not keeping them separated, but also something else. For Kant himself, judgment is in a very important sense synthesis. In Kant’s terms, a judgment is a mental act of synthesis under a logical form, i.e. a form of discursive synthesis (a category), which relates to the synthesis of Mannigfaltigkeit in intuition (Anschauung). What I would like to stress is that this synthetic function of unity is decisive in Kant’s own framework, not least because it is the link to the relations between judgment and consciousness (see Kant 1998, §19, Transcendental Analytics). For Kant a judgment is thus the way given Erkenntnisse (cognitions) are brought to the subjective unity of apperception, or, in other words, to consciousness. Consciousness ‘takes’ things to be a certain way – judgement is for consciousness and by consciousness. So McDowell’s own story about experiencing and judging, which is officially Kantian, and which one might regard as heavily Sellarsian (– or not⁸), goes the following way. Experiencing reveals to us that things are thus and so. He calls this seemings. As we experience, capacities that belong to reason (he speaks of conceptual capacities) are actualized. They are actualized in the experiencing itself. This does not mean that we should think of perceptual experience as ’putting significances together’. All we need to acknowledge is that perceptual experiences are actualizations – not necessarily exercises – of conceptual capacities. There is a potential for discursive activity already there in intuition having its content. Yet, for all that, not all concepts need be at play. Some concepts must, though. Content whose figuring in (such) knowledge is owed to the (further) recognitional capacity need not be part of the experience itself⁹. What matters is that experiencing puts me in a position to know non-inferentially that e.g., what I saw (and I did not then have the concept of, say, cardinal) was a cardinal (imagine you are now looking at a cardinal, which you have seen before not knowing what a cardinal is, and now you see it knowing what a cardinal is).

McDowell’s main point is thus that capacities of reason are actualized even in our unreflective perceptual awareness. He thinks that thinking thus is the best antidote to an intellectualistic conception of human rationality (2009: 271). Now McDowell’s perceptual realism, his view of seeing things as they are, is explicitly meant to be not intellectualistic. He thinks that rationality can be in play in perception without there being any rational supervision or conscious inference by a thinker. This is the explicit intention, and yet one may doubt McDowell’s way is in fact the best antidote possible to an intellectualist conception of human rationality. Hubert Dreyfus, another California philosopher like Searle, and himself a paladin of anti-intellectualism in a conception of the mental, and of perception¹⁰, once criticized McDowell in the context of a well-known debate between the two, by saying that McDowell falls prey to the Myth of the Pervasiveness of the Mental, seeing us as ‘24hour rational animals’, (Dreyfus 2013). In any case, what I want to stress at this point is that McDowell’s view of experience is openly inspired by Kant and that in Kant himself what holds sensible and discursive unity together is the ‘I think’ of apperception. The unity involved in the ‘togetherness’ of sensibility and understanding leads Kant to explore what he calls the originally synthetic unity of apperception, or, in other words, consciousness. Kant’s view of consciousness is a view of consciousness as synthesis. So, since it is unity that McDowell is interested in when he recruits Kant to think about the nature of experience, this is bound to end up leading him to give consciousness an important role when thinking of experience. Like Kant, McDowell believes there is a task for unity here¹¹: the task (for the thinker) of making for the unity of judgment. McDowell sees judgment as ‘the paradigmatic exercise of theoretical rationality’, the making a ‘taking things to be a certain way’ explicit to oneself. As McDowell puts it, judgments are inner analogues to assertions. The capacity to judge is a capacity for spontaneity, for self-determination in the light of reasons recognized as such. Then as he also puts it, a knowledgeable perceptual judgment has its epistemic entitlement in the light of the subject’s experience (McDowell 2009a, b, c, d, e: 257). This is what McDowell is ultimately interested in. Granted, how experience represents things is not under one’s control (McDowell 1994: 11). Yet minimally it must be possible – in view of a particular seeming – to decide whether (or not) to judge that things are a certain way (think of looking at the Müller-Lyer illusion, or looking at a cardinal). This is what McDowell thinks. Such goals and origins of McDowell’s perceptual realism are important if we want to bring out the contrast between his view and Travis’ view. It is crucial that McDowell’s Kantian view of perceptual experience brings along consciousness when thinking of experience, as well as the idea of a task of unity. I will now try to show that both consciousness and the task of unity are conspicuously absent from Travis’ approach to seeing things as they are. As we will see, Travis presents his approach to perceptual experience as Fregean, in contrast with McDowell’s Kantianism.

4 Travis on ‘seeing things as they are’: the silence of the senses and the world’s un-determinacy

In his own realist approach to perception (Travis 2013/2004), Travis replaces McDowell’s Kantian emphasis on the unity of the judgment, as it relates both to consciousness and to the idea of a ‘task’, with Frege’s emphasis on ‘the environment’¹². One does not find the word consciousness in Travis’ formulations. Yet in devising his alternative to McDowell he traces the determinacy he rejects in McDowell’s position to his Kantian appeal to consciousness. By ‘environment’ Travis means what is there to be encountered (by whomever). It could be that we are calculating with natural numbers or looking at the sea in front of our eyes. There is what is there to be encountered. At the same time, with the idea of the silence of the senses, Travis calls attention to (what I will call) the un-determinacy of the world. He thus shifts the focus, in his own approach to perception, from the role of apperception, i.e., of consciousness, which is crucial for Kant’s view of thought-world relations, and is as such inherited by McDowell, to the world and the world’s as it were, un-determinacy ‘in itself’ as we go on perceiving it and thinking about it in more ways, in different ways. In the face of this, what is, according to Travis, ‘seeing things as they are’? Having in mind what we have gone through in Searle and McDowell, I am tempted to use a recent formulation of Travis to bring out the main contrasting point. What is crucial for Travis is to make room for the idea that the world does not auto-articulate (Travis 2021). This is what is hard to capture in thinking about seeing things as they are, as well as in thinking about thinking in general. The ‘ways things are’ is not the world, as it were, articulating itself to let us know how it is, and which ways are these (imagine an Alice-like scenario where a bottle has a tag ‘I am a bottle’ or a piece of gold has a tag saying ‘This is gold and 79 is my atomic number’ – it is not like that). The point is that thought and language have to be in the picture in order to account for any ‘articulation’ of what is there to be encountered (Travis puts much effort in arguing that thought and language ought not to be conflated, but I will not go into that here). Travis’s approach to perception cannot be fully understood without such, as it were, ontological background, as it cannot be understood without his (occasion-sensitive) view of thought-language relations (Travis 2013/2004, 2021). I will concentrate my attention here, though, on this idea that the world does not auto-articulate. Granted, one main, more apparent, target of Travis’ in his approach to perception, especially in earlier writings, were looks and conceptions of looks. More specifically, the target of earlier writings (Travis 2013/2004) was the commitment, shared by many, to the determinacy of looks. Such commitment to the determinacy of looks is present, he claimed, in any of the various forms of representationalism regarding perceptual experience. This is the main target of The silence of the senses (Travis 2013/2004). McDowell was an example of such commitment in Travis 2004, but so were e.g. Peacocke, Searle, McGinn, Harman or Tye. As McDowell himself put it once (McDowell 2013b), for Travis if a rock might look like a crouching animal and also look like a rock, this better not all be the content of the same perceptual experience. In The silence of the senses, a discussion of language (more specifically, an analysis of the vocabulary of looks) supported Travis’ objection to the representationalist’s commitment to determinacy. One main purpose of the analysis was to conclude that there simply is no such thing as a single sense of ‘looks’, a sense which serves the purpose of the representationalist. There simply are many senses of looks or appearances. In particular, there are, in the most recent terminology (Travis 2013) two clearly distinct senses of appearances: ‘perceptual (e.g. ‘visual) appearances’ and ‘conceptual appearances’. Cases such as ‘The upper line looks longer (in the Müller-Lyer illusion)’ and ‘It looks like Macron is going to win the election’ concern, according to Travis, totally different phenomena (namely visual appearances and conceptual appearances, respectively). They should not be conflated. Yet conflation is bound to happen, Travis thinks. It happens namely when we speak of a perceptual experience as a particular seeming. This was the line of argument taken in The silence of the senses. More recently Travis introduced a new Fregean framework for concepts and objects (more specifically for what he calls concepthood and objecthood¹³) as he brought Fregean thoughts (Gedanken, Frege 1918) into the discussion. The work of the Fregean idea of Gedanke is, in Travis’ framework, crucial for thinking of experience and judgment in their relation to the world. Behind Travis opposition to McDowell on perception lies a view of the relation between judgment, thought, and assertion which is very different from McDowell’s, and which is inspired by Frege. These now go together with his former line of criticism based on analysis of ‘looks’. I will not go fully into that now since it brings in different questions regarding language. Suffice it to say here that in Travis’ Fregean framework thoughts comes first and that a Fregean thought (Gedanke) is’ that by which questions of truth arise’ (Frege 1918; Travis 2021). This point is crucial for Travis approach to perception and for his view of any talk of conceptual content. We may regard Travis as addressing some user of the vocabulary of propositional attitudes, someone who, when speaking of conceptual content, is simply taking propositions to be perfectly non-problematic entities, and asking: but what are propositions really? What do you mean by ‘propositions’? (Travis uses ‘thoughts’ (Frege’s Gedanke) instead of ‘propositions’ for reasons he explains (Travis 2013/2004, 2021)). In Travis’ approach to perceptual experience such conception of a thought and of the work of a thought go together with the idea of the silence of the senses. Two elements are crucial in the full picture. First, it is important to bring into view the un-determinacy of the world, of the environment, of what is there to be encountered. The world does not auto-articulate and it can be taken to be many ways. Then, it is important to give a positive account of how we even get something truth-evaluable into the picture when speaking of perceiving ways things are. Without that, we are bound to bring in ‘articulation’ of ways things are, and conditions of satisfaction, unjustifiably, or unjustifiably early (as e.g. Searle does, Travis thinks¹⁴). Travis’ claim is that an environment is needed for there to be accuracy conditions. Thoughts, truth, claims, assertions, and judgments do not float in the air; they need thinkers. By thinkers Travis means not brains qua brains but agents doing the thinking in an ‘environment’. The bottom line is that the Fregean framework does away with the idea that inner perceptual experiencings themselves could have anything like accuracy conditions, conditions of satisfaction. According to Travis only judgments by a thinker in an environment can have such accuracy conditions. What experience does is (just) bring our surroundings into view. An image Travis himself has used is that our senses are like a CCTV; they register. They do not take it e.g., that the flower has five petals, that a robber has just entered the room, or that the sun has set. They are simply registering. Searle would speak of physiology here; Travis speaks of the silence of the senses. Is this the same? I want to argue that it is not. Travis’ (Austinian) idea of the silence of the senses is in fact an alternative to Searle’s contention that all seeing is seeing-as. What Travis wants to get at is this: things ‘as they are’ can be many, many ways (when perceived in these or those circumstances, with these or those instruments, or when thought of anew, as scientific thinking goes on). In fact, we go on knowing more and more about things being as they are, and making further and further claims. The point of the idea of silence of the senses is to emphasize the undeterminacy of what one is presented with, as one is presented with it. The sister-proposal is that the determinateness of e.g. that the flower has five petals or that a robber has just entered the room, or that the sun has set comes in only with our judging-that. Then, naturally, and only then, there are conditions of satisfaction. How this is to be found in Frege is a longer story if we want to make it a matter of interpretation of Frege’s writings. I will simply identify some points in Frege which Travis recruits into his framework. In Logik 1897 (Frege 1983: 149) Frege asks: «But don’t we see that this flower has five petals? One can say that, but then one uses the word ‘see’ not in the sense of a bare experience involving light, but one means by this a connected thought and judgement.» From this Travis takes it that Frege is claiming that seeing-that requires thought and judgment. Frege also claims, in The Thought (Der Gedanke, Frege 1918), that sense impressions do not reveal the outer world to us: «Having visual impressions is certainly necessary for seeing things, but not sufficient. What must still be added in something non-sensible (Nichtsinnliche). And yet this is just what opens up the external world for us; for without this non-sensible something everyone would remain shut up in his inner world. » (Frege 1918: 343). For Frege, this something non-sensible (Nichtsinnliche) is the thought. Travis takes this to mean that only conceptual articulation, which is the job of a thought, opens the outer world for a thinker. Any seeing-that thus requires thought and judgment. Without thought and judgment we would be prisoners of inner space (interestingly, this is a very Kantian idea as well; but Travis wants to attribute it to Frege, as well as to Wittgenstein). The point anyway is that an outer world, an ‘environment’, is required for there to be such a thing as judging truly or falsely. Thus, the business of the philosopher considering perception, or knowledge, in contrast with that of the cognitive scientist, simply requires the business of truth to have entered the picture. The ‘pure business of being true’, to borrow Travis’ term (Travis 2021), has to have set off for there even to be a philosophical problem of perception. Not that there is anything wrong with the scientific business of accounting for cognition or brains. Problems of perception can also be empirical – there is such thing as a science of perception, as practiced by psychologists and neuroscientists. But the problems there are not the philosophical problems of perception. These, when properly philosophical, begin only when thoughts and truth are in the picture and the questions at stake become questions of thought-world relations.

5 Concluding Remarks

One main purpose of this article was to achieve clarity about what is at stake in the formulation of a direct realist position when it comes to speaking of ‘things being as they are’, right in front of our eyes, the things before us. Are things not simply there, in front of me, to be seen as they are? Apparently not, or at least ‘seeing things as they are’ amounts to something very different for Searle, for McDowell and for Travis in their respective approaches to perception. In this article I have identified points where the three authors part ways. So, to recapitulate. First, Searle on the one side, and McDowell and Travis on the other, part ways when it comes to taking perception as it matters to the philosopher to concern, or not to concern, sub-personal events and what goes on causally within our skull, in our brain. This then matters to thinking of perceptual experience as being inside the head or not. A choice regarding naturalism and what we take naturalism to imply methodologically, for philosophy, is involved in this first parting of ways¹⁵. If one goes the way of McDowell and Travis in concentrating on perceptual experience in its relation to the world, and not on what goes on inside the skull, the question then becomes the following. Is consciousness (and the unity of consciousness) involved in there being a particular perceptual experience? McDowell thinks it is, Travis believes it is not. If one then follows Travis in attempting to leave consciousness aside, the question becomes where one stands concerning the determinacy of reality ‘itself’, as it were, in contrast with determinacy in thought and language.

Such questions regarding perception are deep questions, the answers to which in fact extend to the whole of philosophy, and not just to philosophy of perception, and the current disputes within it. Philosophers as diverse as Leibniz, Kant, Husserl or Heidegger grappled with this. I will finish with one remark concerning the last question brought into view, the question regarding determinacy. Commitment to determinacy is crucial for McDowell when it comes to perception. It underlies his representationalism. In his case, it brings into the picture a connection with consciousness, through judgment. Commitment to determinacy is also crucial for Searle. His motto ‘All seeing is seeing-as’, with which I started, brings him to believe that perceptual experience of reality presents reality as (already) determinate. In Searle’s own terms in his 2015 book, ‘reality itself is determinate, as Leibniz told us’ (Searle 2015: 69) (he means, here, to contrast visual experience with a belief or a sentence). Yet Searle seems to forget that Leibniz made an explicit appeal to God’s eye view in putting forward such a claim. For Leibniz, the determinacy of reality is there, fully detailed, in the eye (or mind) of God, and that in fact constrains the way further determinability by human thinking is conceived¹⁶. But what if we cannot help ourselves of anything as that world-unifying God’s eye view, a divine mind which conceptually comprises in itself, in its timeless actuality, the full determinacy of what there is? I want to suggest that Travis goes some way into discussing where we stand, in fact also bringing in Leibniz, and his concept-predicate notion, into the discussions of his Frege-inspired ontology of concepthood and objecthood (Travis 2021, Chapters 7 and 10). Leibniz aside, what is at stake here is how to accommodate perceptual taking of ‘things as they are’ with further thinking of reality – e.g., scientific thinking –, which naturally also concerns things being ‘as they are’. This is a problem of concern to American Realists, e.g., to Wilfrid Sellars, the son, with his theme of the relation between manifest image and scientific image. The challenge for Sellars is how perception, and so the manifest image, stands to the further thinkability of what is there before one, and especially whether perception is to be ‘superseded’ by the ‘scientific image’. I will not go further into this issue, but this is one point I wanted to get to in comparing Searle’s, McDowell’s and Travis’ formulations of perceptual realism. However we choose our path when facing the crossroads identified in this article, one thing is clear: regardless of the intended simplicity of direct realism, there is nothing simple in the idea of seeing things as they are. As soon as we set off thinking about seeing things as they are, we realize there is no single version of what seeing things as they are really is.