A Wassily Kandinsky painting.

A Wassily Kandinsky painting.

Another essay I wrote back in 2008.

Must perception involve the conceptualisation of what is perceived?

‘Perception’ is commonly used to denote the acquisition of knowledge of an external object where the subject’s level of awareness is such that conceptualisation of that object is required. But there is dispute as to whether or not perceptual states include access to pre-conceptual components, that is, whether or not we can have perceptual knowledge of an object without conceptualising it.

Answers to this question fall into two broad traditions, those that view perception as a two-stage process, comprising at first non-conceptual and then conceptual components (both of which we can be aware), and conceptualist models in which we can never experience pre-conceptual components. In this essay two such theses, represented in the accounts of Reid–Dretske and Kant–Strawson [1] respectively, will be contrasted and it will be suggested that there are good cases for both, but the latter is somewhat more plausible, particularly in its view of human understanding.

Both theories agree that there are perceptual experiences and that they are datable episodes culminating a physical process (involving sense organs) from which we gain perceptual knowledge. There is widespread agreement that perceptual experiences have an intentional content (representing the world as being in a particular way) and a phenomenal character (subjective qualities), but much disagreement as to the nature of the content and character.[2] Do we ever experience distinctly non-conceptual content or are we only ever aware of a conceptualised mixture?

The tradition of non-conceptualism begins roughly with the two-stage theory of the direct realist Thomas Reid in the 18th century. Perception consists of sensing and the perceptual thought associated with it. Sensing usually creates a corresponding belief in the agent concerning the qualities of the sensed objects. This belief is part of the judging process which is a distinct act from sensing.[3]

In his interpretation, the contemporary philosopher Fred Dretske emphasises the importance of distinguishing between seeing objects and seeing facts.[4] In the former case, ‘sensory perception’, the information dealt with by the subject is non-conceptual, and in the latter ‘cognitive perception’, it has been conceptualised. Dretske departs from Reid in that whilst he acknowledges these two components as distinct, they needn’t be two separate acts; sense perception of objects may be sequentially prior to cognitive perception but the two are often simultaneous.[5]

Dretske claims that the non-conceptual content of perceptual experiences has a phenomenal character to it, meaning that it directly impacts upon the subject.[6] This content is defined as informational content that does not require the possession of concepts to comprehend it. The criterion for it being a perception is that the information is provided to the cognitive system for calibration and (potential) use in the control of behaviour; this need not involve conceptualisation, or at least not in a linguistic way.[7]

Barry Maund describes another way in which a perceptual experience can have non-conceptual content, that is, when the representation of the object sensed is itself non-conceptual. It signifies the presence of an object, (for example a segment of music alerts us to the presence of a particular arrangement of sound waves) rather than alerting us to the particular properties of the object or its relation to other objects, which would require concepts. He finds this way more plausible because it analyses the phenomenological character of perceptual experience in terms of intrinsic, introspective qualities of experience which is more readily supported by direct evidence than Dretske’s analysis in terms of informational content.[8]

So what makes this theory so appealing? It is intuitively plausible (indeed, it’s likely the default position before we subject the issue to analysis), and it often feels like we are aware of sensory stimulation quite distinctly from any understanding of it. Some such instances are when we experience unusual events or other phenomena when there is a gap between sensing them and knowing what they are. Paul Moser claims that some of these cases will be marked by ‘direct attention attraction,’ this is when we experience shock and/or another response not requiring conceptualisation to events such as an unseen door slamming.[9] Compelling though this argument is, the conceptualist is not convinced that this is evidence of non-conceptual perception.

Conceptualism rejects not only the idea of a non-conceptual stage to perception, but also that our perceptual experiences ever have non-conceptual content. Although we may make more explicit judgements after the original act of perception, perception is always a form of judgement to begin with, [10] as Peter Strawson describes:

“The character of our perceptual experience itself, of our sense experience itself, is thoroughly conditioned by the judgements about the objective world which we are disposed to make when we have this experience; it is, so to speak, thoroughly permeated –saturated, one might say– with the concepts employed in such judgements.”[11]

This is because perception is a biological function which relies upon the processes of receiving and arranging information and conceptualists argue that these processes are best explained by the perceptual system interpreting the information via the medium of concepts. In the higher levels of perception these concepts will correspond to our own human language, but this needn’t be the case for the whole system. Indeed, we needn’t be aware of processing taking place.[12] Conceptualists deny a fortiori that simple sensation can be expressed in the way that non-conceptualists claim it is perceived, nor can it be expressed in simple aesthetic terms; our descriptions of experience at any point in time must be expressed in terms of the concepts of judgement we are naturally disposed towards and constrained to.[13]

Contemporary conceptualism is a radicalisation of Kant’s thesis that perception is a combination of both sensing an object and thinking about it (he held that the processes cooperated, but not that they are inexorably intertwined). Famously he states: “Without sensibility, no object would be given to us, and without understanding none would be thought. Thoughts without content are empty; intuitions without concepts are blind.”[14] In the tradition since Kant the two faculties have not been seen as two components of one integrated act (which Dretske maintains), but as merged from the outset: sensory experiences have no non-conceptual content.[15]

Maund proposes his own hybrid model of a complex sensory and conceptual perceptual experience. In this single act we experience what are distinct sensory and conceptual components but because they are simultaneous and are directed at the same objects we are aware of them only as concept-saturated, whilst they containing some purely sensory content which affects us indirectly.[16] This may in the future provide us with a more viable form of non-conceptualism, but because Maund only mentions it briefly and provides no argument for it, it cannot be properly analysed here. That is not to say that it implausible, indeed, it may be more plausible than Dretske’s position, but the move from simultaneous sensation and conceptualisation to the unconscious efficacy of non-conceptual content is contentious. One might reasonably claim that either this amounts, by extension, to conceptualism, or it is near identical to Dretske’s view but expressed in different language, more sympathetic to conceptualism.

Naturally, there are some substantial objections to conceptualism. One criticism that appears devastating is the possibility for perception without awareness. If we perceive something without noticing it, it is very unlikely to have been conceptualised into human language, and still more unlikely to have been conceptualised in any form, on the basis that concepts are properties that are ascribed by the agent when making a judgement (albeit often passively).[17]

A good case for subliminal perception, based on a study by the psychologist Alberts Libet in 1967, is described here:

“Whereas the receipt and onward transmission of sensory information, initiated by external stimuli, depends upon the classical sensory pathways linking peripheral receptors with their cortical projections, awareness of this sensory traffic –perceptual experience– relies upon sufficient contribution from the ascending fibres of the reticular activating system…. If the reticular system is blocked by surgery or drugs, the arrival of sensory information at the cortex still occurs but the owner of the cortex remains oblivious to the fact!”[18]

Studies by N. F. Dixon (1971 and 1981) have shown that the mind can respond to the meaning of words presented to it in sleep or in other unconscious states, and whilst conscious but below the threshold volume for awareness.[19] An example much closer to our own experiences is attention deficit and change blindness. This is when people clearly see things in front of them but without noticing them (often when something of relative unimportance changes while they are not looking), yet in experiments are able to get the majority of a set of questions right about it by guessing, thus indicating subconscious retention of perceptual information without awareness of receiving it in the first instance.[20]

Whilst we should then accept subliminal perception as real, the most one could hope it proves contrary to conceptualism is that not all perception is conceptualised at a ‘higher’ linguistic level. Even then it seems to only apply to a minority of cases where people are not mentally normal or are otherwise mistaken. But there is also no reason why we should have to adopt a non-conceptualist notion of perceptual judgements in this analysis, if instead we understand conceptualisation to be a process there at the start of perception, one that is largely passive –as suggested above– it needn’t be conscious and so non-conceptual perception has not been proven.

Dretske could then object to this on the grounds that it presumes there is an active perceptual processing system, one that can conceptualise the information received without awareness. For him, the information of the object sensed is sufficiently rich to produce non-conceptual content in perceptual experiences, such that the perceptual processing system acts mainly as a conduit towards higher cognitive processes once this has been achieved; the system is either passive or contributes a minimal amount of inferential reasoning. The problem with this claim is that the constructivist approach, that which promotes active perceptual processing, remains popular, so more evidence is required before a decision can be made as to which is more plausible.[21]

A much better objection to conceptualism is one based on the notion of perceptual learning; Dretske claims that because our ability to recognise certain objects develops as children, this can be seen as evidence that “Sensory perception of objects normally comes before the cognitive perception”.[22] The non-conceptualised sensory information must be the starting-point in early infancy and later still be there underneath conceptual perception. This appears sound, but again there is no reason why infants should not be endowed with faculties of conceptual perception of at least a basic non-linguistic kind, that they are not aware of. Furthermore, Elizabeth Spelke, a psychologist, believes that via a detailed experimental examination of the issue she has produced evidence that infants are innately endowed with a basic concept of objects.[23]

Spelke believes “that the dividing line between perception and thought cannot be determined by introspection. Nor can it be drawn in terms of impenetrability [autonomous in the sense of not being affected by other thoughts or knowledge], learning, passivity, directness or simplicity.” Pure perception results in a simple specification of the layout of what we see without any distinct spatial or temporal boundaries (these are later applied as concepts).[24]

We experience the external world as a layout of physical bodies which endure over time, but this need not be the case in early infancy. Indeed, many have postulated that perception of objects is a learnt skill rather than an innate capacity.[25]

Somehow our visual systems manage to divide up the continuous stream of experience into neatly bounded objects, even when only partly visible.[26] The Gestalt school of psychology challenges this view, positing an unlearned disposition for the perceptual system to organise experience into maximally simple and regular units. This, of course, applies to infants; indeed James and Eleanor Gibson in the 1970s suggested such a system would be highly beneficial- perhaps necessary -for the development of object discernment in infants. But according to Spelke’s observations the Gestalt approach does not provide an adequate explanation, nor does any other traditional model.[27]

This is due to an assumption they share that objects are perceived rather than conceived. She says there is evidence for a central mechanism which organises the surface layout of the perceived area.[28] It seems that the apprehension of objects is a cognitive act brought about by this mechanism, which occurs as soon as the sensory data is collected.

There is evidence that young infants understand “the principle that objects move on spatiotemporally continuous paths.” Studies have shown children recognise the difference between a single moving object which is obscured at a point during its motion and cases where two objects were used rather the one continuous object.[29] “The properties of unity, boundedness, substance, and spatiotemporal continuity appear jointly to constitute an initial object concept.” This central concept then becomes richer as we develop into adulthood.[30]

At the core of her claims is that:

Infants would seem to appreciate two fundamental object properties: substance and spatiotemporal continuity, before they can search for objects that are hidden and even before they can reach for objects that are visible. Prior to the development of most object-directed actions, infants are endowed with mechanisms that carry them beyond the immediately perceivable world.[31]

To conclude, a very good case can be made in favour of Dretske’s interpretation, his revision of non-conceptualism with the claim that the sensing and thinking need not be separate acts makes the theory more plausible because this coheres more with the phenomenological evidence. The most convincing argument for it is probably that on the basis of direct attention attraction, but this is by no means conclusive. Maund’s suggestion of a different kind of non-conceptual content is not much more plausible and as such does little to affect the debate. The conceptualist approach of Kant and Strawson remains slightly more plausible; once we understand its central notion –of all perceptual experience being saturated in concepts– this is a much more plausible theory on an intuitional basis, and is also underpinned by the phenomenology of our perceptual experiences, with which it is largely coherent. But it is Spelke’s work which is decisive, if her evidence is sound from an early age all perception takes place via the two merged faculties. This is not to say that the notion of a distinctly non-conceptual perception is not appealing, indeed this could well take place in subliminal perception. But (especially in regard to the active/passive perceptual processing debate), we will have to wait for more conclusive psychological evidence before we can properly commit ourselves to either position. Overall, it appears more likely that perception must involve conceptualisation than not.


Bergmann, Michael, ‘A Dilemma For Internalism’, In Knowledge and Reality, Crisp, Davis, and Vanderlaan (eds.), (2006), Springer; URL:

Dixon, N. F., ‘Subliminal Perception’, In The Oxford Companion To the Mind, Gregory (Ed.), (1987), Oxford, Oxford University Press, p. 753

Dretske, Fred, [1990], ‘Seeing, Believing and Knowing,’ In Perception, Schwartz, Robert, (ed.), (2004), Oxford, Blackwell Publishing Ltd., pp. 268-286

Dretske, Fred, ‘Perception Without Awareness’, In Perceptual Experience, Gendler, T. S. & Hawthorne, J., (eds.), (2006), Oxford, Oxford University Press, 147-180

Kant, Immanuel, (2003), [1781], Critique of Pure Reason, Meiklejohn, J. M. D., (trans.), Mineola (N.Y.), Dover

Maund, Barry, (2003), Perception, Chesham (UK), Acumen Publishing

Schwartz, Robert, (2004), ‘Introduction To Part IV’, In Perception, Oxford, Blackwell Publishing Ltd., pp. 263-7

Spelke, Elizabeth, [1988], ‘Where Perceiving Ends and Thinking Begins: The Apprehension of Objects In Infancy,’ In Perception, Schwartz, Robert, (ed.), (2004), Oxford, Blackwell Publishing Ltd., pp. 321-336

Strawson, Peter, (1992), Analysis and Metaphysics, Oxford, Oxford University Press


1 This follows the typology of Maund, (2003)

2 Maund, (2003), p. 58

3 Maund, (2003), pp. 58-9

4 Dretske, [1990], pp. 269-70

5 Dretske, [1990], pp. 271-2

6 Maund, (2003), p. 153

7 Maund, (2003), p. 156

8 Maund, (2003), p. 163

9 Bergmann, (2006), p. 29 (footnotes 30 and 31)

10 Maund, (2003), p. 63

11 Strawson, (1992), p.62

12 Maund, (2003), p. 63

13 Maund, (2003), p. 63

14 Kant, [1781], p. 45

15 Maund, (2003), p. 63, with special reference to John McDowell’s theory of conceptualism.

16 Maund, (2003), p. 65

17 Dretske, (2006), p. 147

18 Dixon, (1987), p. 753

19 Dixon, (1987), p. 753

20 Dretske, (2006), pp. 159-67

21 Dretske, [1990], pp. 280-1

22 Dretske, [1990], p. 283

23 Schwartz, (2004), p. 266

24 Schwartz, (2004), p. 266

25 Spelke, [1988], p. 321

26 Spelke, [1988], p. 321

27 Spelke, [1988], p. 322

28 Spelke, [1988], p. 323

29 Spelke, [1988], p. 326

30 Spelke, [1988], p. 330

31 Spelke, [1988], p. 329

Perception and Conceptualisation


2 responses »

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s