2. What is natural science?

Before discussing Merleau-Ponty's position and his arguments for it, some preliminary points need to be clarified, in particular his conception of natural science. In Phenomenology of Perception Merleau-Ponty provides only a sketchy account of this, according to which natural science is primarily a systematic extension of common sense which aims to capture general causal relationships between the objects and properties encountered in experience.Footnote ⁵ In addition, he takes it that natural science exemplifies the atomism which is characteristic of ‘empiricism’ and the mind-independence characteristic of ‘realism’. Merleau-Ponty takes it that these assumptions are rooted in perceptual experience and are indeed appropriate to the circumstances of ordinary life; but, he holds, they are unsatisfactory when conceived as fundamental truths about the world we experience.Footnote ⁶ Merleau-Ponty also claims that once one comes to see the inadequacy of the scientific point of view one will reject the conception of nature implicit in talk of ‘natural science’: ‘We will thus also have to rediscover the natural world and its mode of existence, which does not merge with that of the scientific object’.Footnote ⁷ Merleau-Ponty then devotes chapter 3 of Part II of Phenomenology of Perception to an investigation into ‘The Thing and the Natural World’, whose conclusion is that the natural world is not the mind-independent totality of objects and properties conceived by realist natural science. Instead ‘the numerical determinations of science go back over the outline of a constitution of the world already accomplished prior to them’Footnote ⁸, and the authentic natural world thus retraced is ‘the horizon of all horizons, and the style of all styles, which ensures my experiences have a given, not a willed, unity beneath all the ruptures of my personal and historical life’.Footnote ⁹ A few pages later Merleau-Ponty writes that the world itself ‘in the full sense of the word’ is not an objective totality, for though ‘it is wrapped in objective determinations’, it also has ‘fissures and lacunae through which subjectivities become lodged in it, or, rather, which are subjectivities themselves’.Footnote ¹⁰ This passage suggests that the legitimate aim of science might be to characterise these ‘objective determinations’; but also that this wrapping, being essentially dependent on what is wrapped, is not a complete world. Instead the world is a totality which includes both these subjectivities and their objective wrapping.

Merleau-Ponty returned to the conception of nature in later lectures which have recently been published and translated.Footnote ¹¹ His lectures are interesting but they cover such a diverse range of topics that it is not sensible to deal with them here. It is, however, worth looking back briefly to his earlier book The Structure of Behaviour. Merleau-Ponty here associates natural science with ‘naturalism’ and formulates his claim about the limitations of (natural) science in the following way:

This thesis that ‘perception is not an event of nature’ is central to Merleau-Ponty's position and I shall return to it below. But the context here of this thesis merits further attention. The passage cited above comes at the end of a discussion of ‘Structure in Physics’ whose conclusion is that ‘Laws have meaning only as a means of conceptualizing the perceived world’.Footnote ¹³ Merleau-Ponty reaches this conclusion by arguing that the general structures which physical laws describe are themselves fundamentally ‘forms’ which are dependent upon the perceptual consciousness which exhibits them. It is hard to know what to make of this, and Merleau-Ponty provides no satisfactory examples to substantiate it. On the face of it, however, there is here a sub-Kantian idealism which does not do justice to the abstract complexity of physical law. Think, say, of the tensors which have a central role in general relativity; they are certainly not ‘forms’ that are manifest in the perceived world. Merleau-Ponty does not, I think, repeat this account of natural law in Phenomenology of Perception, but he does maintain his general sympathy for idealist philosophy which is apparent in the conception of the world discussed just above and which runs throughout his critical assessment of natural science.

Before leaving this topic, there is one other complaint that one can, I think, legitimately direct at Merleau-Ponty. Given the strong claims he makes concerning the limits of scientific knowledge, his references to the great French tradition of history and philosophy of science are disappointing. He alludes occasionally to Koyré and Bachelard, but never, so far as I am aware, to Meyerson, Poincaré and Duhem. Equally disappointing are his dismissive remarks about ‘logical positivism’.Footnote ¹⁴ He does not appear to have made any serious attempt to understand the core Vienna circle project of developing and defending a ‘scientific world-conception’. Perhaps he felt that the phenomenological method provides sufficient grounds for dismissing this project without the need to examine it closely. If so, the important contributions of the Vienna circle to the development of 20^th century philosophy show that this judgment was seriously mistaken.

3. ‘Perception is not an event of nature’

In the passage cited at the start from the preface to Phenomenology of Perception, Merleau-Ponty wrote that ‘Scientific perspectives according to which I am a moment of the world are always naïve and hypocritical’. They are hypocritical, he holds, because ‘they always assume, without mentioning it, that other perspective – the perspective of consciousness’. As it stands, this is unpersuasive: the ‘scientist’ will of course recognise that his existence includes his consciousness, and in taking his existence to be ‘a moment of the world’, he must accept that his consciousness is included in this moment of the world's existence. Perhaps Merleau-Ponty can show that this is an untenable view; but it is certainly not so self-evidently untenable as to be such that its denial is naïve or, indeed, hypocritical. It is a substantive task for Merleau-Ponty to show that the scientist's claim is mistaken; and, to do Merleau-Ponty justice, he does indeed attempt this task in ways which are both interesting and distinctive.

The key to his approach is the claim that ‘perception is not an event of nature’ which we encountered just now in the passage cited from The Structure of Behaviour. A similar claim occurs in Phenomenology of Perception: ‘There can be no question of describing it [perception] as one of the facts that happens in the world’.Footnote ¹⁵ Before considering Merleau-Ponty's arguments for this thesis, it is useful to set it alongside other familiar attempts to substantiate the claim that there is something fundamental about human existence which cannot be captured by scientific methods of inquiry. The most venerable line of thought of this kind invokes the conception of a free will, conceived as a capacity to step back from one's natural inclinations and take control of one's life in a way which transcends the natural laws that are the domain of empirical scientific investigation. One familiar way to substantiate this thought is to regard the freedom of a free will as a case of rational self-determination, and thus as drawing on a capacity for rational thought and action which shows how some aspects of human life can be guided by reason rather than just prompted by natural causes. This capacity for rational thought and action is then often regarded as a dimension of human existence which transcends the natural causes investigated by scientific inquiries. But since this capacity draws on a capacity for critical thought which makes the demands of reason evident, it can also be argued that language and the social practices in which it is embedded are the key to the way in which human existence passes beyond the facts of nature. So we find here a third suggestion to be set alongside freedom and reason for the crucial aspect of human existence which escapes scientific inquiry, namely culture.

Turning back now to Merleau-Ponty, it will be apparent that his appeal to perception introduces a novel dimension to this old debate. It is not, of course, that he repudiates the old trinity – freedom, reason and culture. He certainly affirms that the capacity for free action is a crucial aspect of our existence; unlike Sartre, however, he does not take this capacity to be fundamental. The deep truth about us is not that ‘we are condemned to freedom’Footnote ¹⁶ but that ‘we are condemned to meaning’.Footnote ¹⁷ That is, our freedom is made possible by the possibilities inherent in ordinary motivations which are guided by the meanings inherent in the world as we perceive it. In many cases these meanings will be conceptual and expressed in language; but, Merleau-Ponty holds, our capacity for language and conceptual thought is based on a more fundamental, ante-predicative, ‘operative’ intentionality that is manifest in the organisation of the perceived world. Thus although Merleau-Ponty recognises the value of rational thought and action, as manifested in self-conscious reflections of a Cartesian subject, he argues that to take these capacities to be distinctive of human existence is to detach our existence from our lives as embodied agents. And when we attend to our own actions, we find, underneath our rational deliberations, a capacity for unreflective action guided by the meanings inherent in the perceived world.

So Merleau-Ponty's emphasis on perception is intended to underpin traditional accounts of the non-natural aspect of human existence, not to exclude them. Nonetheless his position is original and distinctive. The grounds for it lie in his rejection of Cartesian dualism and the neo-Kantian transcendental idealism that he encountered in the works of the French philosophers he studied, such as Leon Brunschvicg and Pierre Lachièze-Rey, and also in some of Husserl's writings, such as Ideas volume I. If the non-natural aspect of human life is not constituted by the role of an immaterial substance, a Cartesian mind, and does not draw on the transcendent activity of a non-empirical subject, then it has to reside somehow within the capacities of the empirical subject; and where many might point to language as the capacity in question, Merleau-Ponty points to perception, conceived as a capacity for ‘originary knowledge’.Footnote ¹⁸

Yet no sooner is this thesis encountered than questions come crowding in: while freedom, reason and language are all, on the face of it, characteristically, and perhaps distinctively, human capacities, perception is certainly not. But if perception is not ‘an event in nature’, does this imply that animal life in general is not ‘natural’, not amenable to scientific inquiry? So is zoology not a natural science? Or are we to take it that there is a decisive difference between human perception and the perceptions of other animals, such that the former, unlike the latter, is distinctively non-natural? And, anyway, what is it about perception, animal and/or human, that implies that it is fundamentally non-natural?

4. Why is perception non-natural?

In the course of Phenomenology of Perception Merleau-Ponty advances a great variety of considerations to support his account of perception. But because the book is, in large part, an exploration of the significance of our own perceptual experiences, these considerations primarily concern human perception. However, in his first book, The Structure of Behavior, Merleau-Ponty argues that the meaningful organisation of animal perception and behaviour already sets them apart from the existence of things which lack life in a way which precludes the possibility of a scientific account of animal perception in terms of its physical situation.Footnote ¹⁹ He gives a brisk summary of this argument in ‘The Primacy of Perception’:

One thing which is notable here (and in Merleau-Ponty's writings generally) is the absence of any consideration of evolutionary accounts of the functions of the perceptual and motor systems of organisms. While Merleau-Ponty is clearly right to hold that an organism's possession of complex abilities such as perceptual sensitivity and motor skills cannot be produced merely by its exposure to physical stimuli, evolutionary theory suggests that over very long periods of time the development of these abilities can be explained by the selective advantages that have accrued to ancestors of the organism from random changes in the genetic basis of their sensory and motor systems which have modified these capacities in ways that turned out to be beneficial. It may be suggested, however, that this response to Merleau-Ponty's argument is too quick, since his claim was just that ‘perception is not a phenomenon of the order of physical causality’, and one might take it that he is primarily concerned to reject a reductive account of perception according to which it is fully explicable in physical terms. One alternative to this reductionist position is the thesis that capacities such as perception ‘emerge’ over the course of evolution once a system has developed sufficient functional complexity to enable it to process and coordinate information from multiple sensory inputs without the exercise of this capacity being in each particular case simply the effect of the combined activities of its physical sub-systems. This emergentist position is notoriously difficult to characterise and defend, and I shall not attempt this task here. But it is sufficient for my present purpose that the position is not obviously untenable, so that it is worth considering whether it provides a reasonable way of interpreting Merleau-Ponty's account of the status of animal perception which does after all allow for scientific explanations of animal perception, since the methodology of natural science does not carry a commitment to the kind of physicalist reduction that emergentism rejects. What counts in favour of this interpretation is the fact that as well as rejecting reductionist treatments of animal perception Merleau-Ponty rejects vitalist accounts which go beyond emergentism by positing the existence of a special constituent which is characteristic of living organisms, such as Bergson's ‘élan vital’.Footnote ²¹ Like the emergentists, Merleau-Ponty situates his position ‘between’ physicalism and vitalism. However it has to be admitted that his favoured idiom for describing the relationship between the physical and the living is that it is ‘dialectical’.Footnote ²² While I suppose that emergence might be regarded as a kind of materialist dialectic, Merleau-Ponty's use of this idiom seems to be intended to connect with the idealist aspect of his position which I have already noted in his discussion of physical law and which is no part of emergentism. Furthermore, as we have seen, Merleau-Ponty affirms that ‘perception is not an event of nature’Footnote ²³, and although one might say that ‘nature’ is to be understood here as just ‘physical nature’, the general tenor of his discussion is hostile to the possibility of even a non-reductive but scientific account of animal perception. So in the end it seems to me that a straightforward emergentist interpretation of Merleau-Ponty's account of animal perception cannot be sustained, though it is then all the more unclear just what his position amounts to.

What now of human perception? In The Structure of Behavior Merleau-Ponty argues that what is distinctive about human existence is ‘the capacity of orienting oneself in relation to the possible’Footnote ²⁴, and he holds that this separates human life from that of other animals in such a way that human life participates in a kind of dialectic distinct from that involved in animal life.Footnote ²⁵ One thing that is clear here is that it is not at all plausible to propose for this case any extension of the emergentist interpretation of animal perception discussed above. For Merleau-Ponty clearly rejects not just a reductionist account of human perception but, more generally, the conception of human perception as just ‘one of the facts that happens in the world’Footnote ²⁶ which rules out the possibility of any objective, scientific, understanding of perception, including non-reductionist accounts. What, however, are his arguments for this bold claim?

One might have expected him to invoke the central role of mere possibilities in the content of perception which is implied in his emphasis on ‘the capacity of orienting oneself in relation to the possible’. For one can envisage a line of argument which relies on the assumption that only actual causes can have effects, and thus infers that insofar as human perception involves ‘the capacity of orienting oneself in relation to the possible’ it transcends causal determination. However, in his discussions of perception, Merleau-Ponty does not draw on considerations of this kind, though in Phenomenology of Perception he does bring some considerations of this kind into his discussion of the capacity for bodily movement (‘motricité’) in a passage I shall discuss below. Instead here is one line of thought from the Introduction to Phenomenology of Perception:

Before discussing this in detail, it is worth noting that Merleau-Ponty holds that the ‘general effort towards objectification’ which is characteristic of science is itself in part motivated by perception itself. For, as he puts it:

So

Thus the motivation for the kind of understanding characteristic of objective natural science is internal to perception itself. But, Merleau-Ponty holds, this motivation is misguided. Once we manage to suspend this ‘natural attitude’ characteristic of perception and stand back in phenomenological reflection on perception, we come to see that perception has a kind of meaningful subjectivity that cannot be encompassed within the objective framework of natural science.

Why not? If one looks back at the passage cited above, several considerations are apparent:

1. (i) Merleau-Ponty assumes that an objective scientific account of perception requires a reductive explanation which involves ‘a conception of the human organism as a physical system in the presence of stimuli themselves defined by their physico-chemical properties’. He then argues that an account framed in these terms is bound to fail because perception ‘depends on variables such as the biological sense of the situation, which are no longer physical variables, such that the whole escapes the well-known instruments of physico-mathematical analysis and opens onto another kind of intelligibility’.

2. (ii) Along with this assumption of reductionism, there is a further assumption that a scientific approach has to rely on atomic ‘sensations’ conceived as elements of perceptual states; on this view ‘Perception is thereby constructed with states of consciousness as a house is built of stones, and a mental chemistry is imagined that could fuse these materials into a compact whole.’Footnote ³⁰ But, Merleau-Ponty argues, Gestalt theory shows that this atomist approach is mistaken, and that holistic forms such as the figure-ground contrast play a central role in determining perceptual content.

3. (iii) These forms enter into our perceptions as the meaningful structure of experience which enables it to be a way of making sense of things. But, Merleau-Ponty claims, this reference to the role of meaningful structures within perception cannot be accommodated by scientific accounts of perception: ‘objective thought cannot assimilate these phenomena’Footnote ³¹ because ‘one phenomenon triggers another, not through some objective causality, such as the one linking together the events of nature, but rather through the meaning it offers’.Footnote ³² This argument connects with a central aspect of Merleau-Ponty's discussionFootnote ³³, namely a tripartite distinction between causes, reasons, and motives (‘motifs’). Where scientific explanations rely on causal connections and rationalist inferences invoke reasons, in the organisation of perceptual experience the various aspects of experience provide ‘motives’ for making sense of what is being perceived which are neither causes nor reasons. Merleau-Ponty's claim is that there is here a kind of meaningful connection which is fundamental to perception but which cannot be comprehended within an approach which appeals only to objective considerations.

Are these considerations persuasive? The first point concerns the implausibility of reductionist accounts of perception. As I indicated earlier, it is not necessary to take a stand on this issue, since there is no reason to regard non-reductive, emergentist accounts of perception as inherently unscientific. The second point, criticising the atomist presumptions of sensory psychology, may have had legitimate application at the time when Merleau-Ponty was writing, but is now irrelevant. One has only to think of J.J. Gibson's work on sensory systems, such as The Senses Considered as Perceptual Systems, to see that by the 1960s sensory psychology was already detached from the assumptions criticized by Merleau-Ponty.Footnote ³⁴ The third point, however, is the most important one: the issue here is whether one can envisage an explanation of our unreflective capacity for meaningful perception of the world which draws only on natural systems whose functional operations can be understood in causal terms. Merleau-Ponty holds that no such explanation can be sensibly envisaged since causal processes cannot be transformed into capacities which are motivated by meanings. So he treats this capacity as fundamental; it is, he thinks, a kind of intentionality which bears witness to the way in which our embodiment functions as a kind of transcendental subjectivity by giving us the capacity for meaningful perception. However, once one bears in mind both the evolutionary merits one would expect to find built into sub-personal sensory systems and the computational resources available in the brain, it is only to be expected that they should deliver perceptions that make sense of the perceived environment. David Marr's famous account of vision, although speculative and open to criticism in many of its details, shows how scientific inquiries might in principle provide the kind of account which Merleau-Ponty declared to be beyond the reach of science.Footnote ³⁵ More generally, functionalist accounts of content show how systematic causal considerations can be used to assign intentional content to perceptual states which guide actions performed to fulfil desiresFootnote ³⁶. So while Merleau-Ponty identifies plausible requirements on a tenable sensory psychology, once one appreciates the computational resources available in the brain and the potential role of evolution in programming these resources, it is hard to see why an ‘objective’ sensory neuroscience should not be able to accommodate these requirements. Certainly Merleau-Ponty provides no detailed argument to show that this is not possible.

So far the discussion has largely concerned the supposed limitations of a scientific account of perception, which I have argued are misconceived. But what becomes clear as Merleau-Ponty develops his line of thought in Phenomenology of Perception is a more radical thesis which draws on a claim which came up earlier, that although the world ‘in the full sense of the word […] is wrapped in objective determinations’Footnote ³⁷ with which natural science deals, this wrapping covers, or rather, is structured and sustained by ‘subjectivities’ such as our spatiality and temporality, which are integral to perception and exclude the possibility of an objective account of perception.

5. The phenomenal body

Merleau-Ponty's way of addressing this bold claim is slightly oblique. He begins by taking the case of the body and arguing that when we understand properly the nature of our own embodiment we will see that objective natural science does not provide a satisfactory approach to it, and that we require instead an existential analysis which focuses on subjective perspectives that cannot be accommodated within a broader objective, scientific, point of view. Merleau-Ponty initially introduces this attention to the case of the body in part because, he thinks, it is a paradigmatic case in which the scientific approach breaks down –

But, as he goes on to say, the case of the body is not just paradigmatic: it also has a foundational role –

Why does ‘one's own body evade the treatment which it is intended to impose upon it’ in science? Why should we come to treat the ‘objective body’ as just an ‘impoverished image’ of the body ‘as we live it’, the ‘phenomenal body’Footnote ⁴⁰, which is not an object at all but our embodiment? Merleau-Ponty's argument rests on two paradigmatic cases in which, he holds, objective considerations lack the resources to deal with the bodily situation, and in which we need to introduce instead an existential perspective which draws essentially on the subject's experience and way of engaging with his situation. These cases will be familiar, one concerns the phenomenon of the phantom limb experienced by many amputees, the other that of the disability suffered by a patient with a brain injury, Schneider, whose case was studied and written up in detail by the psychologists Gelb and Goldstein. I shall not discuss these cases in any detail; my aim will be just to bring out the points that are central to the use which Merleau-Ponty makes of them so that his general claim about the limitations of objective approaches to the body can be assessed.

6. Space and Time

At the start of Part II of Phenomenology of Perception (‘The Perceived World’) Merleau-Ponty maintains that the fundamental role of perception in providing the framework for knowledge of the world excludes the possibility of providing a comprehensive account of perception within the objective structure of natural science, and this brings in considerations which I have not yet discussed. He provides a memorable metaphorical formulation of this argument at the start of the chapter on ‘Sensing’ (sentir):

The argument here can, I think, be assessed by concentrating on space and time as ‘the horizons opened up by perception’ whose central role in the perceived world entails that perception cannot be just ‘one of the facts that happens in the world’. The argument goes roughly as follows:

1. (a) Any objective, scientific, account of perception must include reference to the objective spatio-temporal relationships between the perceiving subject and the subject's environment.

2. (b) But objective conceptions of space and time are dependent on a subjective, perspectival, understanding of the spatial and temporal relationships which configure the subject's perceived world.

3. (c) So the objective account of perception aimed at in (i) turns out to rest upon the subject's perspectival spatio-temporal configuration of the perceived world; and since this perspectival configuration is precisely that provided by the subject's perceptions, it indicates the ‘lacuna that we are and by which the world itself comes to exist for someone’.

7. Normativity and culture

So far I have concentrated on what I take to be Merleau-Ponty's chief arguments for his critical rejection of the very idea of providing a scientific account of human existence and of the world. In this final section I discuss some further considerations which he advances in support of his position without placing much emphasis upon them.

One type of consideration concerns normativity, and the point here is indeed familiar from current debates. Whereas natural science aims to provide an account not only of what does happen, but also of what is likely to happen or cannot happen or would happen if … etc., it does not of itself lead to conclusions concerning what ought or ought not to happen. So where the exercise of a human ability involves sensitivity to normative requirements of this kind, this fact strongly suggests that the ability is not amenable to a fully scientific, naturalistic, treatment. And if one thinks of the exercise of virtues such as courage, this suggestion does indeed seem to be correct since the identification of these abilities as virtues draws on ethical considerations that are no part of natural science. But what is often disputed when this argument is proposed is whether the exercise of the ability in question really has the kind of normative component which precludes a naturalistic scientific account of its possession and exercise. A famous case in point concerns the ability to carry out arithmetical addition successfully.Footnote ⁷⁰ It is not appropriate to enter into that long-standing debate here, but the debate shows clearly that invoking normativity as a ground for an anti-naturalist conclusion is not a straightforward matter.

Turning back now to Merleau-Ponty, we do indeed find that he holds that there is a kind of normativity in perception:

Hence one might argue that this normativity of perception reveals an aspect of it that transcends scientific explanation. But elsewhere Merleau-Ponty connects this normativity with our capacity for recognizing what we perceive:

So once we take it that a capacity for recognition is one of proper functions of a perceptual system, it is not surprising that the conditions under which this function is best fulfilled with respect to objects of some type become favoured as the ‘optimum’ conditions for perceiving objects of that type. Thus in this case the normativity of perception is implicit in the functional characterisation of perceptual systems which an evolutionary approach would lead one to expect. So this is a clear case in which normativity does not preclude a naturalistic explanation.

But perhaps there are other aspects of perception where an argument of this kind works better. One can suggest, for example, that the relationship between an object's perceptible properties (shape, size, colour etc.) and its potential perceptual profiles is normative: in seeing the black table lamp on my desk in front of me as such I see it as something which should present certain different, but appropriate shapes, sizes, colours etc., when viewed from a range of different positions.Footnote ⁷³ So far as I am aware, Merleau-Ponty does not seek to make his case in quite this way, but his emphasis on the way in which perception finds meaning in what is perceived is perhaps close to this line of thought since meaning and normativity can be closely connected. But rather as in the previous case, more needs to be said to show that the normativity here cannot be accommodated within a naturalistic scientific account of perception. My expectations about how a lamp would appear when viewed from different positions are dispositions acquired over the course of life, and my astonishment if they were not fulfilled in a particular case, is not a matter of normative disappointment, of things not appearing as they should appear, but surprise at my being misled about the visual appearance of something as familiar as a table lamp.

A different line of thought starts from Merleau-Ponty's thesis that perception is ‘originary knowledge’.Footnote ⁷⁴ For if perception is knowledge of this kind, does this not show that perception provides justifications for belief and action which imply that it is inherently normative? One part of the response to this will be an externalist account of justification in terms of reliability: it is the fact that perception is normally a reliable source of evidence concerning the things that are readily perceived which warrants taking it to be a source of knowledge about such things. But the key point here is the distinction between perception as a source of knowledge, albeit a fundamental or ‘originary’, one, and the claim that perception is itself knowledge with the intrinsic normative status this brings. This distinction is not normally important to us; but in this context it is worth drawing since it separates perception as a normally reliable natural way of obtaining evidence about the world from the normative status of knowledge conferred upon this evidence by our normal epistemic practices. Once this distinction is drawn, between perception itself and the epistemic status of the evidence thereby obtained, it is not arbitrary to hold that while our epistemic practices are inherently normative, perception itself is not.

So far I have argued that the appeals to normativity which one can extract, more or less directly, from Merleau-Ponty's discussions of perception do not provide a strong case for holding that there cannot be a satisfactory naturalistic scientific account of perception. But there is a different point which seems to me more substantial, though it brings in many complex and disputable issues. At an early stage of his discussion Merleau-Ponty remarks, concerning a putative scientific account of perception:

There is a serious point here. If a scientific account of perception were supposed to include a reductive programme according to which contents of perception such as ‘the religion whose essence I nevertheless grasp in a hesitation or a reticence, the city whose structure I nevertheless know in the attitude of a city officer or in the style of a monument’ are to be accounted for just by reference to the physical and chemical properties of the stimuli, the position would not be credible. Our understanding of the world, with all its thick historical, personal and cultural references, cannot possibly be captured in a reductive account of this kind. I have in fact already set aside the claim that a reductive programme of this kind is required, since natural science does not rule out an emergentist account of perception. But this point does not altogether defuse the issue raised here. For once one accepts (as I do) that cultural institutions (such as religion, political institutions and architectural style) are not amenable to scientific explanations of the kind provided by the natural sciences because of the role of historically specific traditions, evaluations and practices, it follows that perceptions whose content draws on an understanding of these institutions (as in Merleau-Ponty's examples here) cannot be fully accounted for by reference to the natural sciences.

So we do find here a substantive challenge to the possibility of a complete scientific understanding of perception. But what is also apparent is that this is an issue which is not clearly addressed by Merleau-Ponty in Phenomenology of Perception. The arguments for the thesis that the understanding of cultural institutions is of a different kind from that provided by the natural sciences were developed by the neo-Kantian philosophers, most notably Wilhelm Windelband and Heinrich Rickert, and concern the understanding of historical traditions rather than the phenomenology of perception. Heidegger then combines phenomenological investigations with hermeneutic arguments of this kind; but in this respect Heidegger's work is unique, and is not echoed by Merleau-Ponty. Instead, following Husserl's late writings, Merleau-Ponty concentrates on the thesis of a fundamental ante-predicative operative intentionality that is characteristic of bodily perception, and which he explicitly contrasts with the kind of intentionality that is characteristic of reflective thought and language and which is central to cultural traditions. I have argued that Merleau-Ponty's reasons for holding that our capacity for unreflective bodily perception transcends the limits of natural science are unconvincing. That conclusion is consistent with a recognition that perceptions whose content is informed by cultural traditions transcend natural science, since the intentionality involved here is culturally specific. So there is here a limit to the understanding of perception which natural science can provide. But this limitation does not arise purely from the perceptual process; instead it depends upon the fact that human perception is often informed by culture, which is fundamentally a matter of historical traditions, not scientific laws.