1. What is action-oriented perception?

ZOE DRAYSON

Abstract. Contemporary scientific and philosophical literature on perception often focuses

on the relationship between perception and action, emphasizing the ways in which

perception can be understood as geared towards action or ‘action-oriented’. In this paper

I provide a framework within which to classify approaches to action-oriented perception,

and I highlight important differences between the distinct approaches. I show how talk

of perception as action-oriented can be applied to the evolutionary history of perception,

neural or psychological perceptual mechanisms, the semantic content or phenomenal

character of perceptual states, or to the metaphysical nature of perception. I argue that

there are no straightforward inferences from one kind of action-oriented perception to another.

Using this framework and its insights, I then explore the notion of action-oriented

perceptual representation which plays a key role in some approaches to embodied cognitive

science. I argue that the concept of action-oriented representation proposed by

Clark and Wheeler is less straightforward than it might seem, because it seems to require

both that the mechanisms of perceptual representation are action-oriented and that

the content of these perceptual representations are action-oriented. Given that neither

of these claims can be derived from the other, proponents of action-oriented representation

owe us separate justification for each claim. I will argue that such justifications are

not forthcoming in the literature, and that attempts to reconstruct them run into trouble:

the sorts of arguments offered for the representational mechanisms being action-oriented

seem to undermine the sorts of arguments offered for the representational content being

action-oriented, and vice-versa.

1 Introduction

Contemporary scientific and philosophical literature on perception often focuses on the

relationship between perception and action, emphasizing the ways in which perception

can be understood as geared towards action. Research in psychology, for example,

sometimes characterizes perception as “action-specific” (Ca˜nal-Bruland & van der

Kamp, 2009; Witt, 2011), and talk of perception as being “of affordances for action” is

found in both psychology and philosophy (Richardson, Shockley, Fajen, Riley, & Turvey,

1

2008; Chemero, 2011). Descriptions of perception as “active” are found in vision science

(Whitehead & Ballard, 1990), artificial intelligence (Weyns, Steegmans, & Holvoet,

2004), psychology (Aloimonos, 2013), and philosophy (No¨e, 2004); and as “enactive” in

robotics (Morse & Ziemke, 2007), cognitive science (Froese & Spiers, 2007), and philosophy

(Thompson, 2007). In psychology (Fajen, 2005) and also in sports science (Pijpers,

Oudejans, & Bakker, 2007), perception is described as being “for action”; while the term

“action-oriented” is applied to perception in philosophy (Gallagher & Zahavi, 2014),

computer science (Gora & P., 2014), and neuroscience (Ridderinkhof, 2014).

In what follows, I’ll use the term ‘action-oriented’ as an umbrella-term to include approaches

like these that share the idea of perception as oriented or geared towards action,

in some sense. My first aim in this paper is to provide a framework within which to

classify approaches to action-oriented perception, and to highlight important differences

between distinct approaches.

My framework classifies action-oriented approaches to perception into one of five categories:

1. The evolution of perception as action-oriented: these approaches claim that perception

evolved to guide action, or that perception was selected for its actionoriented

capacities.

2. The mechanisms of perception as action-oriented: these approaches claim that the

mechanisms of perception overlap with, or are closely coupled to, the mechanisms

of action.

3. The contents of perception as action-oriented: these approaches claim that the contents

of perception present the world to the perceiver in terms of possible actions.

4. The phenomenal character of perception as action-oriented: these approaches claim

that a perceptual state is qualitatively experienced as encouraging or demanding a

certain action.

5. The nature of perception as action-oriented: these approaches claim that perception

has a necessary connection to action, that to be a perceiver is essentially to

have certain agentive capacities or skills.

Part of my motivation for providing this framework is to clarify the discussions of actionoriented

perception across philosophy and the sciences. I will show that each approach

within the framework is logically independent of the others: none of the approaches can

be derived from the others without further justificatory assumptions. We should thus not

conclude from the fact that perception is action-oriented in one respect that it is actionoriented

in any of the other respects.

This paper also has a second aim, which is to use this framework to explore the role of

action-oriented perception in embodied cognitive science. In particular, I will focus on

the claim that perceptual representations can be action-oriented. I will show that Clark

(1997) and Wheeler (2005) put forward a concept of action-oriented perceptual repre-

2

sentations which is committed to both the representational mechanisms of perception

being action-oriented, and to their contents being action oriented. But given that neither

of these claims can be derived from the other, proponents of action-oriented representation

owe us separate justification for each claim. I will argue that such justifications are

not forthcoming in the literature, and that attempts to reconstruct them run into trouble:

the sorts of arguments offered for the representational mechanisms being action-oriented

seem to undermine the sorts of arguments offered for the representational content being

action-oriented, and vice-versa.

In Sections 2-6, I introduce each of the distinct respects in which perception can be said

to be action-oriented. Section 7 explores the use of action-oriented perceptual representations

in embodied cognitive science.

2 The evolution of perception as action-oriented

There is an evolutionary sense in which perception can be described as action-oriented,

that is often expressed with the claim that perception is “for action”. Proponents of this

view emphasise that our perceptual capacities evolved to guide our bodily interactions

with the world, specifically those interactions which increased our adaptive fitness. Talking

about visual perception, for example, Briscoe claims that:

“[f]rom an biological or evolutionary standpoint, it is reasonable to think

that vision is for action, that its preeminent biological function is to adapt

an animal’s bodily movements to the properties of the environment that it

inhabits” (Briscoe, 2014, p. 202, my italics).

On one interpretation of such claims, the idea that perception is “for action” in an evolutionary

sense seems trivially true, and it’s not clear whether any believer in evolution

would deny it. But the evolutionary claim is often put forward in a stronger way: the

assumption is sometimes that perception evolved directly to guide action and not (or at

least not directly) to present the perceiver with an action-neutral description of the objective

world. If perception evolved for its action-guiding properties, so this line of thought

goes, then it did not evolve to enable us to have beliefs about how the world is, independently

of our own actions. Proponents of this view either conclude that we do not

perceive the objective world in an action-neutral way, or that if we do, this is not the

primary function of our perceptual systems: perception did not evolve in order to provide

inputs to the human capacity to think and reason about the world. Representative

versions of this view can be found in the works of philosophers like Patricia Churchland

and Kathleen Akins:

“looked at from an evolutionary point of view, the principle function of nervous

systems is to get the body parts where they should be in order that the

organism may survive. [. . . ] Truth, whatever that is, definitely takes the

hindmost.” (Churchland, 1987, p. 548)

3

“evolution will favor sensory solutions that package the information in efficient

and quickly accessible formats, in ways that match the particular physical

form of the motor system, its motor tasks, and hence informational requirements.

[. . . ] the elegant solutions that evolution eventually selects need

not involve any straightforward (to our eyes) ‘veridical’ encoding of sensory

information.” (Akins, 1996, p. 353)

Notice that this involves a strong commitment to a particular evolutionary story. Proponents

of this view have to rule out the possibility that the development of amodal

cognition could have been more adaptive than a faster but less flexible system in which

specific sensory inputs drove specific motor outputs. Some scientists claim exactly this,

arguing that game theory demonstrates that selection pressures would make objective

representations of the world an unlikely outcome (Hoffman, Singh, & Mark, 2013), but

this is often assumed rather than argued for. Notice that even if one accepts that evolution

selected for action-guiding perception, this is consistent with the claim that evolution

also selected for the kind of perception that can yield reasoning capacities and abstract

thought. It might be the case, in other words, that we have two forms of perception. This

is a possibility raised by the dual-visual system theory of perception, discussed in the

next section.

3 The mechanisms of perception as action-oriented

According to a traditional picture of the mind, perceiving and acting are distinct mechanisms

which are separate from each other and from our thinking mechanisms: perception

provides input for thought, and action emerges as the output of thought. (This is what

Susan Hurley (1998) terms the ‘classical sandwich’ picture of the mind.) This picture

is challenged by empirical work which suggests that the mechanisms of perceiving and

acting are closely intertwined, overlapping, or even co-constituting. If this is true, then

there is a sense in which the mechanisms of perception can be action-oriented regardless

of their evolutionary history.

One example of this comes from the literature on so-called ‘mirror neurons’: in macaque

monkeys, neurons known to control hand and mouth movements fire both when the monkey

is manipulating an object (e.g. reaching for a piece of food) and when the monkey

is watching someone perform a similar manipulation. In creatures like humans, where

single-cell recordings are not possible, there is similar data showing that areas of the

brain associated with movement are active during perceptual processing: one and the

same neural mechanism seems to be involved in both perception and action. Perception

is action-oriented in the sense that a perceptual mechanism seems to overlap with a

mechanism for action.

In visual psychology, Milner and Goodale (1995) have advanced the ‘dual visual systems’

hypothesis, which concerns the way that visual perception builds up information

4

about the world. They have demonstrated that sensory input to the visual system of primate

brains can be processed by two independent pathways. The processes in the ventral

pathway reflect the traditional picture of perception as the input to thought: they produce

conscious perceptual states that we can categorize, memorize, and combine with

thoughts to produce a broad range of actions. The processes in the dorsal pathway, however,

do not seem to be the inputs to thought: these visual inputs instead lead only to the

specific motor outputs involved in reaching and grasping objects with our hands. The

dorsal processes seem to be action-oriented in the sense that visual input results in the

appropriate motor output without the intermediary of conscious thought.

This sort of ‘close coupling’ between sensory input and motor output is also found in

robotics and artificial intelligence. The traditional view of perception and action as distinct

mechanisms separated by central thought processes creates engineering problems

for designers of intelligent systems by causing bottlenecks to occur: the sorts of thinking

required to update plans and amend instructions slows down the system’s ability to

respond to sensory stimuli. Roboticists like Rodney Brooks produced systems that could

interact with their environments in real-time, by cutting out the central ‘mind’ focusing

instead on using specific input sensors to generate specific outputs. Brooks (1991) proposes

that arranging these sensory-motor couplings in ‘layers’ in the appropriate way,

surprisingly intelligent behavior can be produced. One layer might govern general locomotion,

such that a robot will move around at random until it hits an obstacle. Another

layer might then take over and turn the robot until the way is clear, before control reverts

to the first layer. A third layer might sense red light and keep the robot on track to reach

the light, thus overriding the first layer’s random movement. Each sensory process in

these robotic architectures is thus action-oriented in the sense that that each layer has

its own sensors which operate exclusively for one kind of behavior. There is no amodal

cognition or control: the communication between the layers is minimal, and amounts to

just switching each other on and off. Such architectures can be used to create systems

that display a remarkable amount of seemingly intelligent flexible behavior from purely

reflex-like action-oriented sensory-motor couplings. It remains an empirical question

whether such architectures can simulate higher-level behaviours, such as deciding between

actions, without the addition of central thought mechanisms.

The three examples I’ve considered – mirror neurons, the dorsal visual stream, and

sensory-motor architectures – are examples of ways in which the mechanisms of perception

could be said to be action-oriented. Notice that claims about action-oriented

perceptual mechanisms are independent from the previously considered claim that perception

is action-oriented in the sense of having evolved to guide action. First, consider

the robotic architectures just described. The fact that an optimal engineering solution

uses action-oriented perceptual mechanisms (in the sense of reflex-like couplings between

sensory inputs and motor outputs) doesn’t tell us that human perceptual systems

actually evolved such couplings: evolutionary solutions are often the satisfactory but suboptimal

ones, rather than the optimal solutions. Second, while the dorsal visual stream’s

use of sensory input to directly guide action seems to support the evolutionary claim,

it does so only with respect to a certain subset of perceptual processing: the guiding

5

of fine-grained motor control, rather than action more generally. And the ventral visual

stream seems to have evolved to generate the kind of perceptual states that can input into

thought, which is exactly what the proponents of the stronger evolutionary claims want to

deny. Third, despite many claims about the evolutionary purpose mirror neurons might

have served, it’s not clear that we can draw any conclusions about the action-oriented nature

of perceptual evolution from the empirical data. Cecilia Heyes (2009), for example,

has persuasively argued that mirror neurons are a byproduct of our general capacity for

associative learning and not the result of evolutionary adaption.

4 The contents of perception as action-oriented

Many of the claims about action-oriented perception in philosophy concern the contents

of perception: how the world is presented to us in experience. It is traditionally assumed

that perception presents the world to us in a way that is entirely neutral with respect to

the actions one might perform: while the contents of beliefs might present the world to us

in terms of how we can act on it (presenting food as edible or stairs as climbable, for example),

such contents are not available in perception, according to the traditional view of

perceptual content as action-neutral. Proponents of action-oriented perceptual content,

on the other hand, propose that action-involving properties like edibility or climbability

can be presented to us in the contents of perception: “we see objects as edible, and

do not just believe that they are” (Nanay, 2012, p. 430). Following Gibson (1979), we

can use the term ‘affordance’ to describe properties like edibility and climbability, and

characterize the action-oriented view of perceptual content as the claim that we perceive

affordances. Action-oriented views of perceptual content come in different strengths,

depending on how they characterize the idea of affordances. To perceive the world as

affording a certain action might, on a weak reading, mean that we perceive the possibilities

for action: the sorts of actions we are capable of performing. A stronger reading

would claim that action-oriented perceptual content presents the world to us in a way that

solicits or encourages us to act in a certain way; and an even stronger reading views affordance

properties in perceptual content as mandating or demanding a particular action.

Action-oriented perceptual content need not be consciously experienced by the perceiver,

but there is a way of extending the idea of action-oriented perceptual content to include

the phenomenal character of perception, which I’ll discuss in the next section.

As I am using the notion of perceptual content here in a weak way, according to which

talk of perceptual content is talk of the world as presenting itself to the perceiver in experience.

Such talk is not committed to the stronger claim that perceptual experience

fundamentally consists in the subject perceptually representing their environment as being

a certain way. The weaker notion of perceptual content is therefore not committed to

representationalism about perception, but is compatible with at least some forms of relationalism

(Siegel, 2014). Furthermore, proponents of action-oriented perceptual content

might conceive of content as propositional or non-propositional; as structured conceptually

or non-conceptually; as Russellian or Fregean, or in terms of possible worlds; and

6

as environment-dependent or independent.

It is important to emphasise that since action-oriented perceptual content is not committed

to this content being carried by a particular neural or psychological mechanism, then

a fortiori it is not committed to the existence of action-oriented perceptual mechanisms.

This will be further discussed later in Section 7.

5 The phenomenal character of perception as action-oriented

Philosophers draw a distinction between the content of perception and its phenomenal

character. To talk of perceptual content is to talk of how perception presents the world;

whereas to talk of the phenomenal character of perception is to talk of the qualitative

properties of the perceptual state, or what it feels like to undergo the perceptual experience.

If we assume that perceptual states need not be consciously experienced, then

is possible for a perceptual state with the sort of action-oriented content outlined above

to lack phenomenal character altogether. And where a perceptual state has phenomenal

character and action-oriented content, that action-oriented content might be “nonsoliciting”

(Siegel, 2014): one could consciously perceive something as affording action

without experiencing the motivation to perform that action. Similarly, Prosser (2011)

proposes that the phenomenal character of a perceptual state correlates with its actionoriented

perceptual content, but doesn’t claim that the phenomenal character itself is

action-oriented.

But it looks like there could be cases where the phenomenal character of a perceptual

state is itself action-oriented. Siegel (2014), for example, claims that perceiving what

the environment affords can sometimes be accompanied by a felt quality of solicitation,

where the perceptual experience has a qualitative feel of inviting or prompting us to act. If

we think of this as action-oriented phenomenal character, then it looks like phenomenal

character might be action-oriented to differing degrees: some perceptual experiences

might not just feel like invitations or prompts, but more like urges that motivate us to

act. Siegel (2014) proposes that this particular subset of perceptual states with actionoriented

phenomenal character have a “feeling of answerability” and are experienced as

mandates for action.

Notice that how one understands the relation between action-oriented content and actionoriented

phenomenal character will depend on one’s attitude more generally to the relation

between perceptual content and phenomenal character. A proponent of strong intentionalism

about phenomenal character, for example, would presumably claim that any

action-oriented aspects of phenomenal character reduce to the action-oriented properties

of the perceptual content. At the other end of the spectrum, one might think that phenomenal

character is entirely independent of perceptual content, and thus that it is possible to

have two perceptual experiences with the same content but distinct characters.

7

6 The nature of perception as action-oriented

I have suggested that claims about perception being action-oriented can be understood

as making a variety of different claims about the evolution, mechanisms, content, or

character of perception. But there remains a further question about the modal strength

of the claims being made. Should we understand any of these claims as proposing that

perception is necessarily action-oriented in any of these senses? And if so, what kind of

necessity is involved?

I take it that most of those making action-oriented claims concerning the evolution of perception

or the mechanisms of perception are making largely empirical claims about features

of perception in the actual world. But at least some proponents of action-oriented

perception seem to be making modal claims, which can be interpreted in at least two

ways. On one hand, the necessity at play might be nomological necessity. To say that perception

is necessarily action-oriented in the nomological sense is not to say that perception

is action-oriented in all possible worlds, but rather to say that the laws of nature in the

actual world make it physically impossible for perception not to be action-oriented in the

actual world. Alternatively, the necessity in question might be metaphysical necessity,

concerning the nature of perception across possible worlds. To posit a metaphysically

necessary connection between perception and action is to claim that there is no possible

world in which perception is not action-oriented in some appropriate sense.

The claims made by some proponents of sensorimotor theories of perception, such as

Alva No¨e (2004) and Susan Hurley (1998), seem to suggest that there is a constitutive

dependence of perception on our capacity for action: that to be a perceiver is necessarily

to be an agent. It is not clear what kind of necessity is involved, but their reliance

on empirical evidence suggests that they are primarily concerned with making a case

for nomological necessity. The case for action being metaphysically necessary for perception

is perhaps found in Schellenberg (2010) exploration of the relationship between

perception, self-location, and spatial know-how. Her key claim, that perception requires

the capacity to know what it would be to act in relation to objects, seems to rely on our

intuitions about how perception could be in possible worlds rather than how it is in the

actual world.

Notice that even where claims of nomological necessity are made and supported by empirical

evidence about the mechanisms of perception, there is not direct entailment from

action-oriented perceptual mechanisms to the claim that perception is necessarily actionoriented.

We cannot simply read off claims of nomological necessity from empirical

facts (Callender, 2011). And claiming that perception is necessarily action-oriented in

the sense that Schellenberg intends does not entail that either perceptual content or phenomenal

character is action-oriented. In fact, Schellenberg proposes that perception’s dependence

on action capacities is what allows us to perceive more than just the relational

or perspectival properties associated with affordances: she claims that the action-oriented

nature of perception can account for our access to the intrinsic, observer-independent

properties of the objects perceived.

8

7 What about action-oriented representation?

I’ve differentiated several different approaches to perception that fall under the umbrellaterm

‘action-oriented’, and demonstrated that each approach makes claims that are independent

from the others. There is no entailment from one variety of action-oriented

perception to another without the addition of further premises.

While I’ve talked about the mechanisms of perception and the content of perception, I

have said little about the matter of perceptual representations: content-bearing internal

states of the perceiver. In fact, many people working in action-oriented perception want

to play down the role of representations. This is because in cognitive science, representational

mechanisms are traditionally posited to account for the flexibility of intelligent

human behaviour: our capacity to respond differently to similar stimuli, and similarly

to different stimuli, in a way that can’t be easily accounted for in terms of reflex-like

responses. Positing internal representations allows cognitive science to say that the same

stimulus can be represented in different ways: the same extensional content can be carried

by different representational ‘vehicles’ which play distinct causal roles within the

cognitive mechanism.1 These vehicles are just internal representations individuated by

the non-semantic properties in virtue of which they are causally efficacious: the formal,

functional, or physical properties to which the cognitive mechanisms are sensitive. Thus

two representations are tokens of the same vehicle type when they are treated similarly

by the cognitive processes at play (Shea, 2007).

Traditionally, cognitive science suggests that these representational mechanisms are genuinely

cognitive in the sense of being non-perceptual: they are distinct from the mechanisms

of perception, and they are what allow the same perceptual input to result in

different action outputs. Proponents of embodied cognitive science, however, deny that

explanations of flexible and context-specific behaviour require this sort of mediating representational

mechanism. They argue instead that if perceptual mechanisms are actionoriented

in the sense discussed in Section 3, then these sorts of coupled mechanisms can

account for the behaviour.

But not everyone agrees that the insights from action-oriented approaches to perception

should lead to the rejection of internal representational states. There are many within

embodied cognitive science who allow that there are internal representations, but deny

that these are the sort of amodal cognitive representations in a central cognitive architecture.

Andy Clark (1997), for example, argues that behaviour is not mediated by “actionindependent

inner states; ones which require additional cognitive operations to drive

appropriate behavior”, but rather by action-oriented representations which are “poised

between pure control structures and passive representations of external reality” (Clark,

1997, p. 49). A similar approach is put forward by Michael Wheeler, who understands

1For naturalistic reasons, cognitive science is generally committed to extensional content.

9

the idea of action-oriented perceptual representations as “special-purpose adaptive couplings”

that are “tailored to a particular behavior” (Wheeler, 2005, p. 196). Both Clark

and Wheeler propose that the sorts of action-oriented mechanisms of perception discussed

in Section 3 are best understood as representational mechanisms: as containing

perceptual states which are both causally efficacious and semantically evaluable. They

argue that these perceptual representations themselves, and not just the mechanisms in

which they occur, should be understood as action-oriented because in virtue of the way

that they are causally coupled to certain motor processes.

To claim that a mechanism is representational is to be committed to vehicles of content.

But this does not commit one to any particular views about the nature of that content:

what it is or how it is determined, evaluated, or structured. (We might agree that a list of

items is a representation, while disagreeing over whether it is a description or an instruction,

for example; we might believe a drawing to be a map without knowing what it is

of; we might know that someone’s utterances are linguistic without knowing what they

mean.) To understand an internal state as a representational vehicle does not, therefore,

entail a particular view of its content. A fortiori, it does not entail that its content is actionoriented

in the sense of presenting the world in action-relevant ways. The reverse is also

true, as already discussed in Section 4: action-oriented perceptual content is neutral with

regard to the mechanisms of perception, and therefore compatible with a range of cognitive

architectures. Action-oriented perceptual content does not entail action-oriented

perceptual mechanisms, and a fortiori does not entail that there are action-oriented vehicles

of representation bearing the action-oriented perceptual content.

Interestingly, however, both Clark and Wheeler seem to assume that their action-oriented

vehicles of perceptual representation have action-oriented perceptual content. Clark

claims that action-oriented perceptual representations “simultaneously describe aspects

of the world and prescribe possible actions”, with the result that “to thus know the world

is at once to know what possibilities it affords for action and intervention” (Clark, 1997,

p. 49), while Wheeler claims that action-oriented perceptual states “represent the world

in terms of specifications for possible actions” with “bearer-relative content” (Wheeler,

2005, p. 196). Their action-oriented representations, in other words, seem to be examples

of action-oriented perceptual mechanisms with action-oriented perceptual contents.

Neither Clark nor Wheeler explain why they think that action-oriented vehicles will have

action-oriented contents. Any attempt to reconstruct their arguments will, I suggest, encounter

the following problem: reasons to think of perceptual vehicles as action-oriented

seem to undermine our reasons to think of perceptual contents as action oriented, and

vice-versa.

To see this, consider the role played by representational vehicles. Recall that vehicles

are just internal representations individuated by the non-semantic properties in virtue

of which they are causally efficacious within the representational mechanisms. Vehicles

provide a naturalistic way of understanding how the same extensional content can

play distinct cognitive roles, such as when the same heavenly body is thought about in

10

different ways according to the time of day it is observed. The extensional content is

the same, but the representational vehicles have distinct causal properties in each case

and thus interact differently with other states of the mechanism. If we claim, as Clark

and Wheeler do, that representational vehicles should be individuated according to their

action-guiding properties, then this enables us to show how the same extensional content

(e.g. a mountain) can be represented as climbable in one context and unclimbable in

another context: distinct vehicles, with different causal relations to action, can carry the

same extensional content. But if this is the role of representational vehicles, to provide

a naturalistic means of accounting for modes of presentation, then it removes the need

to think of the content itself as action-oriented. The perceptual content can be actionneutral

because the relevant action-guiding information is supplied by the perceptual

mechanisms.

One might, however, have independent motivation for thinking that affordance properties

are part of the perceptual content itself: we perceive the climbability of the mountain, for

example, rather than perceptually representing the mountain as climbable. But if we

don’t need to perceptually represent the mountain as being as certain way, then it’s not

clear what motivation we’d have for positing vehicles of representation, rather than nonrepresentational

mechanisms. This, I take it, is the essence of Gibson’s (1979) view of

perception. And even if we had reason to posit representational mechanisms, we’d then

need a further argument for individuating the representational vehicles according to their

action-guiding properties.

In summary, the arguments for taking perceptual vehicles to be action-oriented seem to

undermine the arguments for taking perceptual contents to be action-oriented, and the

arguments for taking perceptual contents to be action-oriented seem to undermine the

arguments for taking perceptual vehicles to be action-oriented. None of this rules out

that there may be arguments from the action-oriented status of either vehicles or contents

to the other, but it suggests that the onus is on proponents of such views to provide

them.

11

Bibliography

Akins, K. (1996). Of sensory systems and the “aboutness” of mental states. Journal of

Philosophy, 93(7), 337–372.

(2013). In Y. Aloimonos (Ed.), Visual navigation: from biological systems to unmanned

ground vehicles. Psychology Press.

Briscoe, R. (2014). Spatial Content and Motoric Significance. Avant: Trends in Interdisciplinary

Studies, 2, 199–216.

Brooks, R. (1991). Intelligence without representation. Artificial Intelligence, 47, 139–

159.

Callender, C. (2011). Philosophy of Science and Metaphysics. In S. French & J. Saatsi

(Eds.), Continuum Companion to the Philosophy of Science (pp. 33–54). Continuum.

Ca˜nal-Bruland, R. & van der Kamp, J. (2009). Action goals influence action-specific

perception. Psychonomic Bulletin & Review, 16(6), 1100–1105.

Chemero, A. (2011). Radical embodied cognitive science. MIT Press.

Churchland, P. S. (1987). Epistemology in the age of neuroscience. Journal of Philosophy,

84(10), 546–553.

Clark, A. (1997). Being There: Putting Brain, Body, and World Together Again. MIT

Press.

Fajen, B. R. (2005). Perceiving possibilities for action: On the necessity of calibration

and perceptual learning for the visual guidance of action. Perception, 34(6), 717–

740.

Froese, T. & Spiers, A. (2007). Toward a phenomenological pragmatics of enactive perception.

Gallagher, S. & Zahavi, D. (2014). Primal impression and enactive perception. Subjective

Time: The Philosophy, Psychology, and Neuroscience of Temporality, 83–99.

Gibson, J. J. (1979). The Ecological Approach to Visual Perception. Houghton Mifflin.

Gora, P. & P., W. (2014). Adaptive System for Intelligent Traffic Management in Smart

Cities. In International Conference on Active Media Technology (pp. 525–536).

Springer International Publishing.

Heyes, C. (2009). Where do mirror neurons come from? Neuroscience and Biobehavioral

Reviews.

12

Hoffman, D. D., Singh, M., & Mark, J. (2013). Does evolution favor true perceptions? In

Proc. SPIE 8651, Human Vision and Electronic Imaging XVIII (Vol. 865104).

Hurley, S. (1998). Consciousness in Action. Harvard University Press.

Milner, M. A., A.D. & Goodale. (1995). The Visual Brain in Action. Oxford University

Press.

Morse, A. & Ziemke, T. (2007). Cognitive robotics, enactive perception, and learning in

the real world. Cognitive Science.

Nanay, B. (2012). Action-oriented Perception. European Journal of Philosophy, 20, 430–

446.

No¨e, A. (2004). Action in perception. MIT Press.

Pijpers, J. R., Oudejans, R. R., & Bakker, F. C. (2007). Changes in the perception of

action possibilities while climbing to fatigue on a climbing wall. Journal of sports

sciences, 25(1), 97–110.

Prosser, S. (2011). Affordances and Phenomenal Character in Spatial Perception. Philosophical

Review, 120(4), 475–513.

Richardson, M. J., Shockley, K., Fajen, B. R., Riley, M. A., & Turvey, M. T. (2008).

Ecological psychology: Six principles for an embodied-embedded approach to behavior.

In Handbook of cognitive science: An embodied approach (pp. 161–187).

Ridderinkhof, K. R. (2014). Neurocognitive mechanisms of perception–action coordination:

A review and theoretical integration. Neuroscience & Biobehavioral Reviews,

46, 3–29.

Schellenberg, S. (2010). Perceptual Experience and the Capacity to Act. In M. M. Gangopadhay

& F. Spicer (Eds.), Perception, Action, and Consciousness (pp. 145–

159). Oxford University Press.

Shea, N. (2007). Content and Its vehicles in connectionist systems. Mind and Language,

22(3), 246–269.

Siegel, S. (2014). Affordances and the Contents of Perception. In B. Brogaard (Ed.),

Does Perception Have Content? (pp. 39–76). Oxford University Press.

Thompson, E. (2007). Mind in life: Biology, phenomenology, and the sciences of mind.

Harvard University Press.

Weyns, D., Steegmans, E., & Holvoet, T. (2004). Towards active perception in situated

multi-agent systems. Applied Artificial Intelligence, 18(9–10), 867–883.

Wheeler, M. (2005). Reconstructing the cognitive world: the next step. MIT Press.

Whitehead, S. D. & Ballard, D. H. (1990). Active perception and reinforcement learning.

Neural Computation, 2(4), 409–419.

Witt, J. K. (2011). Action’s effect on perception. Current Directions in Psychological

Science, 20(3), 201–206.

Author biography. Zoe Drayson is Assistant Professor of Philosophy at the University

of California, Davis. She previously held positions at the Australian National University

and the University of Stirling, and gained her PhD from the University of Bristol. She

works in philosophy of mind and cognitive science, with a focus on the nature of psy-

13

chological explanations across philosophy and the mind sciences. Her published work

explores the distinction between personal and subpersonal approaches to the mind; the

commitments of embodied and extended cognitive science; the relation between perception,

cognition, and action; the metaphysics and epistemology of explanation; and

various issue surrounding the nature of mental representation.

14

:: بازدید از این مطلب : 131

|

امتیاز مطلب : 0

|

تعداد امتیازدهندگان : 0

|

مجموع امتیاز : 0