(Article originally published in 2002 Journal of Consciousness Studies 9(12), 65-88:  journal web page:



Abstract:  This article supports the proposition that, if a judgment about the aesthetic merits of an artistic object can take into account and thereby be influenced by the particular quality of the object, through gestalt experiences evoked by the object, then we have free will.  It argues that it is probable that such a judgment can indeed take into account and be influenced by the particular quality of the object through gestalt experiences evoked by it, so as to make it probable that we do have free will.  The proposition is supported by reference to two basic tricks apparently involved in conscious processes, which I call the qualia trick and the chunking trick; and it is suggested that these tricks make possible and indeed probable the existence of a third trick, which I call the selection trick. 



In this article, I’m going to support a stark and somewhat surprising proposition – that if a judgment about the aesthetic merits of an artistic object can take into account and thereby be influenced by the particular quality of the object, through gestalt experiences evoked by the object, then we have free will.  And I mean real free will, not just the compatibilist appearance of free will.

I will also argue that it is probable that such a judgment can indeed take into account and be influenced by the particular quality of the object through gestalt experiences evoked by it, so as to make it probable that we do have free will.  And I will consider how my arguments apply to other kinds of judgments - theoretical, moral, and practical.

          I will support this by reference to two basic tricks apparently involved in conscious processes, which I call the qualia trick and the chunking trick; and I will suggest that these tricks make possible and indeed probable the existence of a third trick, which I call the selection trick.  So far as we know, none of these tricks are performed by computers, as presently understood.  The first two of them are the subject of some scientific investigation, and there are some ideas as to what kinds of brain processes are required if they are to be performed.  The third is barely studied, and indeed is disregarded by most scientists; and few people have any ideas as to what it might involve.  It is this third trick that would be the basis for free will.


1  Laws and Rules

Part of my argument relates to the role of laws and rules in the making of aesthetic and other judgments, so it is necessary to say a little about laws and rules.

There appear to be regularities in the way that processes unfold in the world, and it is common to regard these regularities as reflecting laws that govern or constrain what happens in the world, some at least of which are called ‘laws of nature’.  These laws have not yet been, and may never be, formulated with precision by scientists, and those laws that have been formulated are often expressed as relations between variables rather than in terms of what causes what; but it is generally assumed that there exist laws, to which the formulations of science approximate, and which are causal in the sense that, in their totality, they constrain what changes occur in the world, either deterministically or by restricting these changes to some range or spectrum of alternatives.  I will use the word ‘laws’ generally to refer to any laws or rules that constrain what happens in the world, whether or not they would be called laws of nature.

I think it is a matter of convenience whether one considers these laws as being in some sense distinct from the properties of the features of the world with which they engage, such as mass, electric charge, spatial separation, direction, motion, etc., or as being inherent in the features themselves or in the matter, energy, fields, etc. which display these features.  But however one regards the laws, it seems clear that one feature they must have is generality:  in order that they be laws, they must be such as to engage with features that the different states of affairs or objects or processes or events to which they apply may have in common. 

And just as the laws must have generality, so also must the features of the world with which the laws engage.  Very broadly, laws are to the effect that when X occurs Y will occur; and both X and Y must then be features that a number of different states or objects or processes or events may have in common.

          Generality here does not altogether exclude uniqueness.  To the extent that X and Y incorporate variables that are quantitative and susceptible to mathematical analysis, the uniqueness of particular quantities involved in particular cases of X and Y will not preclude engagement with constraining laws.  The archetypal laws of nature, namely the laws of physics, apply to quantities of physical properties such as mass, electric charge, etc., and they apply to any such quantities and any combinations of such properties; and so it does not matter that a combination is unprecedented and will never occur again, so long as it can be given quantitative expression.  The laws prescribe the quantitative effect of quantitative variations.  Even something unique and as complex as a particular state of the weather may be considered as constituted by a limited number of common features of varying quantities (temperature, pressure, etc.), so that its uniqueness does not preclude the engagement of that state with constraining laws.  In relation to the weather, there is no reason to think that there are any other features that do not engage with constraining laws and yet do contribute to the determination of how the system changes:  I will argue that it is otherwise in relation to conscious experiences.

          To the extent that any X and Y are not quantitative, they must, in so far as they engage with constraining laws as presently understood, be the same whenever they occur:  any variations that are not quantitative will not engage with the law ‘when X occurs Y will occur’.  And the rarer and more complex X and Y are, the less likely it is that there will be a law of nature that engages with them.  At the extreme, if X is a feature that occurs just once in the history of the universe, then a constraint to the effect that when X occurs Y will occur would only have the one occasion for its application and so would not properly be called a law.

          Thus for example, unless there were more than one Big Bang, a constraint that applied only in the circumstances of the Big Bang could not properly be called a law, because it would have only the one occasion for its application.  But in so far as the conditions that existed in the Big Bang are quantitative, they may have engaged with the same constraining laws that apply at other times:  it would not matter that the particular quantities involved have never existed and will never exist at any other time.  Even if, for example, heat and density of mass at this time were infinite, the same laws as apply to finite values of these quantities may have applied; and calculation of the result of application of such laws to infinite values might be possible.

          In this discussion, I am making two assumptions, which I should now make explicit.

          First, I am assuming that there is indeed a valid distinction between what is quantitative and what is not quantitative.  I take it that, in so far as any feature (and by this I intend to include any aspect of any feature and any difference from any other feature) can be adequately described by numbers or other mathematical objects, to that extent the feature is quantitative; and in so far as any feature cannot be adequately so described, to that extent it is not quantitative, and may be called qualitative.  It is beyond the scope of this article to analyse this distinction in detail:  it is sufficient for my purposes if the distinction is understood in general terms, and if it is accepted that some features of the world cannot adequately be described by numbers or other mathematical objects that can be incorporated as variables in constraining rules.  I do not suggest a narrow limit to what is quantitative:  as will be seen, I accept that variations in tones and intensities of visual experiences of colour may be quantitative, in that such variations may perhaps be adequately described by numbers or other mathematical objects.  However, I do assume, for example, that differences such as that between a visual experience and a feeling of pain, or that between a visual experience of a colour and a visual experience of a line or edge, are not merely quantitative:  I assume that these differences cannot be adequately described by numbers or other mathematical objects.  It may in fact be that all non-quantitative features are features of conscious experiences, but my argument does not require any assumption about this.

          Second, I am assuming that constraining laws have some minimum simplicity and broadness of application; so that, if a law is to the effect that when X occurs Y will occur, then, in so far as X and Y are not quantitative, they must to some minimum extent be simple and commonly occurring.  Again, I do not suggest a narrow limit on this.  I certainly do not assume that all constraining laws must be or derive from laws of physics.  However, I do assume, for example, that a rich visual experience is too complex and rarely occurring for it to engage as a gestalt whole with constraining laws, at least except in so far as it may be quantitative; so that if such an experience does engage with constraining laws, it must be through those of its features that are quantitative, or sufficiently simple and commonly occurring to so engage, or both.  Features of that kind I will call general features or features having law-engaging generality.  Other features, that are neither quantitative nor sufficiently simple and common to engage with constraining laws, I will call special features.  Special features may or may not be had in common by different states, objects or processes; but whether they are or not, they will not engage with constraining laws because they do not have law-engaging generality.

          I think that both these assumptions are modest.  Although some people make wide claims about what is quantitative, I think everyone accepts that there are at least some basic entities that cannot be adequately described or distinguished from each other by numbers or other mathematical objects.  And my second assumption is widely made, albeit often tacitly, in science and philosophy.

What I am suggesting then is that states, objects, processes and events of the world combine with constraining laws, so as to contribute to the determination of what changes occur in the world, only through their general features that can engage with such laws.  And in so far as features of states, objects, processes and events are neither quantitative nor of some minimum simplicity and frequency of occurrence, they do not have the generality that is required for engagement with constraining laws so as to contribute to the determination of what changes occur in the world.

          One view of the way causation operates in the world is that there are succeeding states of affairs, and that constraining laws operate on each of them so as to determine either a single line of development from it, or alternatively a spectrum of lines of development one of which occurs at random within probability parameters determined by the laws.1   On this view, although the whole of any state of affairs, in its full particularity, is important in the causal process as constituting initial conditions for subsequent development, the actual line or spectrum of lines of development from that state of affairs is determined entirely by the combination of the relevant constraining laws and the general features of these initial conditions with which they engage.  Thus, if a state of affairs has a feature or combination of features that does not engage with any law of nature because it does not have generality in the sense explained above, then that feature or combination of features cannot be efficacious in determining what changes in the world are brought about in accordance with the laws:  the efficacy of the state of affairs in that respect must be through those of its features that do have generality in the sense explained above.


2  The Game of Life and Computation

The role of laws in the unfolding of events can be illustrated by reference to the Game of Life.  This game was devised in about 1970 by John Conway, a Cambridge mathematician.  Its rules can be stated shortly:

Life occurs on a virtual [and potentially infinite] checkerboard.  The squares are called cells.  They are in one of two states:  alive or dead.  Each cell has eight possible neighbours, the cells of which touch its sides or corners.

If a cell on the checkerboard is alive, it will survive in the next time step (or generation) if there are either two or three neighbours also alive.  It will die of overcrowding if there are more than three live neighbours, and it will die of exposure if there are fewer than two.

If a cell on the checkerboard is dead, it will remain dead unless exactly three of its eight neighbours are alive.  In that case, the cell will be ‘born’ in the next generation (Levy 1993, 52).

Given an initial configuration of live and dead cells, everything that happens in the game is determined unequivocally by these basic rules, which may be considered as analogous to the laws of physics.

          Features on a larger scale than the one-cell-and-eight-neighbours scale dealt with by the rules may be causally efficacious in the sense that they are part of initial conditions that are modified by the rules engaging with constituent elements of these features, so as to produce further larger-scale features; but they do not themselves engage with the rules so as to bring about changes.  This is true even of features that appear to unfold in accordance with larger-scale rules, because their causal efficacy in bringing about changes is only through their properties at the one-cell-and-eight-neighbours scale.   For example, there is a five-cell pattern called a glider which, after four generations in which no other live cells are encountered, results in an identical pattern displaced diagonally by one cell.



The rules of the game dictate that the state of any cell in any generation is wholly determined by the state of that cell and the eight adjoining cells in the preceding generation; so the rules do not in fact engage with the glider pattern as such.  Of course, the glider pattern is itself causally efficacious in the sense that it is modified by the rules engaging with constituent elements of the pattern so as to produce further glider patterns; but there is nothing in the game itself that recognises or responds to a glider as such, either as cause or effect.  An outside observer may recognise a glider, and may construct a rule that a glider will continually move diagonally across the checkerboard, progressing by one cell every four generations, unless it encounters any other live cells.  This chunking of the five live cells of the glider, and their adjacent dead cells, provides an observer with a useful description of what is happening, and may help the observer to understand and predict the game’s unfolding; but the glider pattern itself is not causally efficacious in bringing about any changes, because efficacious causation of this kind is entirely at the one-cell-and-eight-neighbours scale.  The glider pattern does not as a whole engage with the relevant laws.

          It might be said that the rules of the Game of Life could be restated in such a way that the glider pattern engaged with the rules of the Game, and that according to this restatement the glider as a whole would be causally efficacious in bringing about changes.  That argument has some plausibility, because the glider is a simple and commonly-occurring pattern, and therefore the sort of feature to which general laws could plausibly apply:  it would have far less plausibility in relation to complex and rarely-occurring patterns of the kind to which general laws would not plausibly apply.  But in any event, any rule in this restatement which concerned a larger-scale pattern such as a glider would have to be qualified so as to ensure that in no situation did it conflict with the two basic rules of the Game, and these two rules would have to be included in any event to cover situations not covered by rules concerning larger-scale patterns.  Thus it would seem that, in any rational enquiry for relevant rules, it would run counter to Occam’s razor to postulate a rule applying specially to gliders; although I will consider why this may not be so later in this section.



The fact is that, however the rules were stated, their true effect would be that, if the five live cells of a glider were those numbered 1 to 5 in the above diagram, then the live state of cell 1 would have absolutely nothing to do with the state, in the next generation, of the cells numbered 3, 4 and 5, because it is not adjacent to them.  So in that sense at least, even in relation to such a simple and common feature as a glider, the glider pattern as a whole does not engage with the rules of the Game so as to contribute to the determination of what changes occur in accordance with the rules; although all patterns that occur do contribute to the determination of what happens by providing initial conditions which are modified by the operation of the laws on those features that do engage with the laws.  And just as there is nothing in the Game that recognises and responds to a pattern such as a glider, there is nothing in the Game that groups or chunks the cells constituting such patterns into wholes or ‘gestalts’ so that they can as wholes play some role in the Game.



It would be different if, instead of just restating the rules of the Game of Life, one could modify them so that, for example, looking at the above diagram, the state of cell A in the next generation was not wholly determined by its eight neighbour cells, but was affected by the complete glider pattern.  Suppose, for example, that cell A is dead and that the rules of the modified game are such that (1) if just three of A’s eight adjacent cells are alive, then it may or may not be alive in the next generation; and (2) in that event, it will be alive if and only if those three live cells are themselves part of a glider.  Subject to what I say below about programs, it is only by that kind of modification to the rules that a ‘gestalt’ such as the glider pattern would truly engage with the relevant laws and thus be truly efficacious in bringing about changes, and that it could make sense to say that there is something in the Game that chunks the cells constituting that pattern into a whole so that it can as a whole contribute to the way the Game unfolds.

The kind of information-processing carried out by computers as presently understood can readily be explained in terms of rules or laws analogous to the rules of the Game of Life.  A request for certain information or for a solution to a certain problem may be entered into a computer, giving rise to a state of the computer which, we may suppose, is unique and unprecedented.  The constitution of the computer and its program is such that constraining laws engage with certain general features of the state in question, and of succeeding states, in such a way as to give rise ultimately to a state of the computer by which the answer (or perhaps both the request and the answer together) is displayed.

          Subject to what I say below, there is no reason to think that the computer or its program grasps the request or the answer as a whole, except to the extent that the entire process I have described could be considered as amounting to such grasping.  The transition from the state representing the request to the state representing the answer is mediated entirely by the operation of general constraining laws that engage with general features of the initial and succeeding states.  This would mean that any novelty or uniqueness in objects, processes, or states of affairs dealt with by the request, and thus reflected in the initial state, is not taken into account by the computer, except to the extent that this novelty is reflected in numbers or other quantities with which the laws engage:  the initial conditions would be novel, but the line or lines of development from them would be constrained by the constraining laws engaging only with their general properties.  The state representing the answer (or the request and the answer together) may thus reliably indicate that the earlier novel state occurred, if it can be interpreted as doing so.  But for a computing machine, as presently understood, such interpretation of the answer would at best be by means of further processes of the same general kind.

          It has in fact been shown that the Game of Life, on a sufficiently large scale and given sufficient time, can operate as a computing machine, capable of solving any computable problem (see Poundstone 1987, 197-217).  Thus, there would be a configuration of the Game that would constitute it as a computing machine, and a modified configuration that would represent the machine with a request for information or for a solution to a problem.  The progress of the Game in accordance with its two rules would ultimately give rise to a further configuration representing the machine with the information or solution.

As we saw earlier, the two rules of the Game operate entirely at one-cell-and-eight-neighbours level, and there is no need and no role in the unfolding of the Game for the ‘chunking’ of any more cells than this; and even if one postulated that the rules of the Game included larger-scale rules engaging with chunked cells or patterns like gliders, these rules could only be such as to engage with commonly occurring general patterns and thereby to give the same result as given by the two rules operating at the one-cell-and-eight-neighbours level.  They could not conceivably be considered as engaging with whole configurations representing particular requests or answers.

So again it seems clear that there could be no grasping by the Game of the request or the answer, unless the entire process from request to answer is taken to be such grasping.  And while the entire process might conceivably be considered as amounting to a grasping of the request, it could not conceivably be considered as amounting to a grasping of the answer:  that would surely require at least some further process.  And if the answer is not grasped, it seems implausible to suggest that the request has been.  (Readers may detect here an echo of Searle’s Chinese room.)

          If one found in the Game some process whereby there was, in the course of the progression from the request to the answer, some kind of chunking of patterns so that they could be grasped as a whole, this would be exceedingly odd.  There is simply no role for such chunking or grasping in the unfolding of the Game, which, as we have seen, develops entirely in accordance with the operation of rules at the one-cell-and-eight-neighbours level.

          What I have said so far about the Game of Life and computation is subject to an important qualification.  There is a sense in which chunking does occur in computations, whether they be performed by a computer or by the Game of Life:  computing machines run in accordance with programs that operate at levels higher than the level of basic physical causation in computers or the one-cell-and-eight-neighbours level of the Game of Life; and elements of the programs must be represented by and engage with larger-scale features of the system.  Furthermore, the same program can be realised and run on different systems, and so to that extent the programs may be considered as independent of the basic features and the basic rules.  In that sense, the rules of the program may fairly be considered as having their own efficacy and significance, and certainly as not being eliminable by Occam’s razor.  The point can be taken further:  it may be that at larger scales again, the Game or a computing machine changes over time in accordance with statistical rules, that may not be derivable from the basic small-scale rules and that accordingly may also be considered to be independent of those basic rules.  These various rules additional to the basic rules might be called emergent rules.

          However, in the Game of Life and in a computer as presently understood, even emergent rules would never require or permit anything to happen otherwise than in accordance with the basic rules; and the emergent rules themselves must engage with general features of the relevant system, as discussed in the previous section.  In such systems, any features not having law-engaging generality would not contribute to the determination of what happens, except as being aspects of initial conditions that are modified by the operation of laws on general features.


3  Two Tricks of Consciousness:  Qualia and Chunking

When I think about our conscious processes in comparison with computational processes such as those just discussed, it strikes me that consciousness seems to involve two very important tricks, that have no place in computational processes.

First, there is the qualia trick, the apparent trick of associating certain types of neural processes with certain types of subjective qualitative experiences or feelings.  It seems, for example, that there are types of patterns of neural activity that correspond with visual experiences of various kinds, and other quite different types of patterns of neural activity that correspond with feelings of pain.  Thus, a pattern of type A could correspond with a green patch in a certain part of the visual field, a pattern of type N could correspond with a vertical edge in another part of the visual field, and so on.  This type of correspondence is the subject of considerable scientific investigation, and there has been some success in associating types of processes with types of conscious experience; but as yet no understanding of the function of this correspondence and no idea how to reproduce it artificially.

This first trick on its own may not seem to have great causal significance:  just as the type of neural pattern corresponding to a visual experience of green has general features that can engage with constraining laws (like the state of individual cells and their eight neighbour cells in the Game of Life), so also the corresponding visual experience of green may have general features that could be appropriate to engage with such laws (like a glider in the Game of Life).  I will suggest that it is in combination with the second trick that the qualia trick could have enormous significance.

So second, there is the chunking trick.  Consciousness also has the apparent trick of bringing together a multitude of general qualia into a particular global experience that is had by a particular subject.  This trick is associated with what is called the binding problem of consciousness:  how it is that information carried by processes in distinct parts of the brain is brought together into a unified conscious experience.  This problem is also the subject of considerable scientific investigation, and again there has been some success in identifying types of processes that give rise to this ‘binding’, but no understanding of the function of this binding and no idea how to reproduce it artificially.

Like the first trick, this one does not require anything other than the operation of general laws on general features:  the constraining laws seem to be such that, whenever a type of process occurs in a region of a person’s brain (perhaps a pattern including the coordinated 40Hz oscillations of the activity of relevant neurons, which is sometimes said to be part of the solution to the binding problem), the types of qualia associated with other types of processes also occurring in that region are chunked together into a whole gestalt experience of the person.

          One possible objection to my identification of two tricks is to the effect that there can be no qualia without a subject to experience them, and there can be no subject without chunking; and thus there cannot be two separate tricks.  There is force in that objection, but I think it is likely that there are (1) types of neural processes that give rise to chunking (such as processes like 40Hz oscillations), and (2) different types of neural processes that give rise to qualia, either by themselves or in combination with the processes that give rise to chunking.  If the combination of processes (1) and (2) is required to give rise to qualia, then the processes in (2) could be considered as giving rise to potential qualia that become actual qualia when chunked into a whole experience of a subject; and with this modification the identification of two tricks could stand.

The global experience resulting from the two tricks will display general features with which general laws could engage, but will also display particular and often unique combinations of these feature.  A visual experience, for example, will display features of colours, shapes, patterns, and so on, that could have law-engaging generality, and also particular combinations of these features; and we seem to grasp the particular combinations as wholes, as unified gestalts.  I have assumed that such a visual experience is not sufficiently simple or commonly occurring to engage as a whole with constraining laws, except in so far as it may be quantitative; and I suggest that, because a visual experience is a combination of qualia, it cannot be adequately described by numbers or other mathematical objects and thus is not quantitative.  Therefore, I contend, such an experience cannot as whole engage with constraining laws.

            The following table shows this schematically, in relation to a grossly over-simplified2 account of seeing a work of art:                 


Brain processes

A:  pattern a in region x

B:  pattern b in region y


N:  pattern n in region z


R:  pattern r in regions

x + y + … + z + …        


Conscious processes

Aq:  red patch in bottom left of visual field

Bq:  white patch in bottom right of visual field

Nq:  vertical edge in top right of visual field

(Aq+Bq+…+Nq+…):  gestalt experience in which all

visual elements are chunked


Apart from the bottom-right item, every item in the table can plausibly have law-engaging generality, so that it could affect what happens by engaging with general constraining laws.  Further, every step or correspondence between items on the left of the table and items on the right could be in accordance with general constraining laws.  If, as I think likely, the relationship between items on the left and items on the right is a kind of identity (in the sense that both are aspects of the one process), rather than item on the left causing item on the right, then all items on the right could have effects in accordance with general laws through their correspondence with items on the left:  the items on the right would in that respect be like gliders in the Game of Life, that is, features that might be regarded as themselves engaging with laws, but are more satisfactorily regarded as having effects through lower-level features with which they correspond.  But in addition, items on the right, apart from the bottom one, could have effects in accordance with general laws through their own law-engaging generality.  Thus there could plausibly be a constraining law to what happens if Aq occurs (if there is a red patch in the bottom left of the visual field), and it could be that various tones and intensities of red also have the generality required for engagement with constraining laws, either because each of these tones and intensities is sufficiently simple and commonly occurring to engage with constraining laws, or because such variations are merely quantitative.  Indeed, such a qualia/law combination could be efficacious quite independently of the efficacy of the neural correlates if, for example, the relationship between the neural correlates and the qualia were not a kind of identity, and particularly if the neural correlates caused the qualia in some circumstances and not in others.

But I contend that this could not be the case for the bottom item on the right.3  There could not plausibly be a law as to what happens when the gestalt experience (Aq+Bq+…+Nq+…) occurs:  such an experience has gestalt features which are neither quantitative nor sufficiently simple and commonly occurring to engage with constraining laws, and which therefore do not have law-engaging generality.  Thus, any engagement with constraining laws by such an experience must be:

(1) through its constituent qualia; and/or

(2) through other general features of the experience; and/or

(3) through its neural correlates.

And that means it must be an engagement that does not operate through any features produced by the chunking of the qualia which do not have law-engaging generality.  Of course, it may be that engagement through general features does in fact account for all the effects that the gestalt experience has, so that the gestalt experience merely supervenes as an epiphenomenon on these features; but I will argue to the contrary, in particular because this would mean that the chunking trick has no plausible function.

          I am not saying, as writers like Donald Davidson (1970) have done, that there can be no rigorous application of general constraining laws to any mental events.  I accept that, to the extent that even qualitative features of mental events have generality, they could engage with and be rigorously governed by such laws.  Rather, I say that what precludes rigorous application of constraining laws to gestalt experiences is the combined effect of the qualia trick and the chunking trick, giving rise to particular combinations of qualia which, as gestalt wholes, lack the generality necessary for engagement with constraining laws.

It is reasonable to believe that chunking has a function.  It has been said that its function is to integrate information, and it may be that chunking of qualia is associated with physical processes that do integrate information in a way appropriate to computational operations; but this would not explain the function of the chunking of qualia, in the absence of an account of how chunked combinations of qualia are used. It is plain that computers as presently understood could not be programmed to use chunked combinations of qualia.

It could be suggested that the function of chunking is no more than to display general features additional to the general features that were combined to create the chunked experience, so that these additional general features can then engage with constraining laws (cf. category (2) above).  For example, the chunking of two large dots side-by-side, a central vertical line below them, a horizontal line below that, and a surrounding circle, may disclose a general pattern suggestive of a human face.  But general patterns of this nature can be detected without chunking, by a system set up to do so:  our computers can do it, and it seems clear that most of our own basic recognition of types like human faces is carried out pre-consciously, by evolution-selected algorithms, without the assistance of either the qualia trick or the chunking trick.  And plainly, chunking does also produce special features that do not have law-engaging generality, and it is reasonable to look for a function for the production of those features.

     What I will suggest is that the function of these two tricks is, by means of the application of general laws to general features, to constitute a particular experience with particular qualities, had by a particular subject, who can then select a development from a spectrum of possible developments left open by the operation of general laws on general features.  This selection is neither determined by rules nor random, but is based on the subject’s grasp of particular wholes; and it is facilitated by the way in which combinations of qualia present enormous amounts of information simply, graphically, comprehensively, and with salience given to important features.  I say more about this in Section 5.

          Finally in this section, I should explain how the argument of this section stands with my earlier assertion that, where you have a law that when X occurs Y will occur, both X and Y must have law-engaging generality.  It could be said that my argument presupposes that when R occurs (Aq+Bq+…+Nq+…) occurs, yet (Aq+Bq+…+Nq+…) does not have law-engaging generality.  However, while there is a law presupposed in the last line of my table, and it operates whenever R occurs, what it then prescribes is not that (Aq+Bq+…+Nq+…) occurs but that operation C (for ‘chunking’) occurs, whereby whatever qualia are associated with the patterns of neural activity within the region in which pattern r occurs, are chunked into a gestalt experience.  It is operation C that has law-engaging generality, and this operation gives rise to the chunked combinations of qualia that do not have law-engaging generality.


4  Aesthetic judgments

So now I come to consider judgments about the aesthetic merits of artistic objects, and the role of particular gestalts in such judgments.  When I refer to judgments about the aesthetic merits of an object, I am referring both to judgments made by the artist in creating the object, and also to judgments made by other persons in appraising or simply enjoying the object; and I am referring to judgments ranging from bare judgments of approval or disapproval through to elaborately considered and articulated assessments of the object’s merits.  The artistic object in question may be of any kind – literary, musical, visual, etc.  The argument is perhaps clearest in relation to the creation and early appreciation of ground-breaking works such as Picasso’s Les Demoiselles d’Avignon and Wagner’s Tristan und Isolde, but it applies to artistic works and gestalts of all kinds.

          I will begin by discussing in a very general way what kinds of conscious and unconscious processes are involved in a person’s making an aesthetic judgment.  Undoubtedly there are extensive and highly complex unconscious processes that must take place in the person’s brain if any such judgment is to occur.

(1)     There are the complex pre-conscious processes that are necessary in order that one have a conscious experience of the work.  In the case of the visual arts, there are all the pre-conscious processes that produce the appearance of a viewed object or scene, not merely processing its configuration but also coordinating data from both eyes so as to give depth, and compensating for eye movement so as to give stability.  These pre-conscious processes also include the processes that enable one’s memory and understanding of the world to contribute to the recognition and interpretation of what is seen, and thus in turn to how it looks.

(2)     There are also pre-conscious processes that give rise to the emotions that seem to be involved in aesthetic judgments.  The person will to some extent become conscious of these emotions, but the conscious feelings result from and express much pre-conscious activity.

(3)     When a person sets about making a judgment concerning what the person is experiencing (seeing, hearing, etc.) and feeling, there is plainly a great deal of non-conscious information-processing underlying whatever conscious thoughts may occur.

          In addition to these unconscious processes, there are also subjective conscious processes that seem be involved in the making of a judgment.  In saying this, I am not asserting that these subjective conscious processes are wholly distinct from objective brain processes:  indeed, it seems clear that conscious processes must have neural correlates, and as noted earlier, I believe that the conscious processes and their neural correlates are best considered as two aspects of the same processes.

(1)     When a judgment is made about an artistic work, the work is generally, for some time at least, present to consciousness.  One doesn’t appraise a painting without actually seeing it and consciously taking it in, and one doesn’t appraise a musical performance without hearing it and attending to it.

(2)     Such judgments generally involve some emotions that are consciously felt.  I have said that there must be unconscious processes that give rise to emotions involved in the judgment, but I would maintain that at least some of these emotions are present to consciousness as the judgment is made.

(3)     If the judgment is a carefully considered and articulated judgment, there are generally conscious thoughts that take place in which one formulates and assesses reasons for the judgment.  Whether these conscious thoughts add anything to their neural correlates and other non-conscious processes, and if so how, is a moot point, on which I hope this article sheds some light; but I think it is undeniable that such conscious thoughts do occur.

          Many of the processes I have mentioned as taking place in aesthetic judgments engage with causal laws and are thereby constrained by them.  It is reasonable to accept that pre-conscious processes underlying the experience of seeing or hearing, and underlying emotional feelings, are computation-like processes that proceed in accordance with the constraints of laws which engage with the general features of the objects, states of affairs and processes involved.  Further, the relationship between the neural correlates of elements of the conscious experiences, and these conscious elements themselves, seems to be law-like, so that general features of the elements of conscious experiences correspond in a regular way with general features of the neural correlates.  As we have seen, it seems that there are types of patterns of activity in the brain that correspond with types of patterns of conscious experience.  It is reasonable to think there would also be general features of conscious experiences or their neural correlates that engage with genetic and/or learned predispositions in the brain so as to contribute to emotions and judgment.  For example, types of patterns in the visual field suggestive of human faces, or of types of landscapes beneficial to survival, whether they are consciously recognised as such or not, would presumably contribute to emotional responses in law-like ways.

          Thus laws can engage with all kinds of general features of pre-conscious brain processes, and of experiences and their neural correlates; and they can engage with every feature necessary to give rise to and uniquely specify, define and/or determine every possible conscious experience.  Yet the conscious experiences themselves are particular and often unique – unprecedented and never-to-be-repeated.  And although all the general features that give rise to and specify and together determine a particular experience can engage with constraining laws, the particular experience itself and its special features cannot do so, because they lack the necessary generality.

          Consider a painting like Picasso’s Les Demoiselles d’Avignon.  As an object, it is unique, particular, and unprecedented.   It is constituted by basic constituents (atoms, molecules) in a configuration that can be expressed in mathematical terms representing the relationships between those constituents; and on a larger scale, it has general features, such as basic colours, edges, and shapes, with locations and relationships that can be expressed in mathematical terms.  On the basis of these general features, passable reproductions can be produced.  More significantly for our purposes, we can have a visual experience of the painting, constructed by law-governed processes operating on general features of the painting and of our brains, so that, through the first two tricks of consciousness, general features are chunked into whole gestalts that we can experience and grasp.  We can thus experience and grasp unique qualities of the painting, such as the look of particular faces, and the painting’s overall appearance, just because these qualities are captured by the combination of the qualia trick and the chunking trick.  But I suggest that the experience of these unique qualities cannot engage with general laws; so either the experience is not taken into account and does not thereby influence our appraisal of the painting, or it is taken into account and thereby influences our appraisal otherwise than by engagement with general constraining laws.

          When this painting was first created by Picasso, and when it was appraised by early viewers, we may suppose that both Picasso and the viewers were influenced by emotions triggered unconsciously by law-governed processes.  But it also seems likely that they did experience gestalts that expressed qualities of the painting and various parts of it; and the utter novelty of these gestalts confirms that any account taken of the gestalts could not be entirely the result of general rules engaging with them, whether they be constraining laws applying to physical processes, or evolution-selected algorithms, or aesthetic laws determining what counts as a meritorious painting.  (I am not of course suggesting that the historical context in which the painting was created is unimportant:  plainly, the readiness of Picasso and early viewers to respond positively to the painting was due in part to that context.)

          To take another example:  consider the first time Richard Wagner grasped the opening bars of the Prelude to Tristan und Isolde, when he conceived them in his imagination or perhaps when he played them on his piano for the first time.  We may assume that there were general features of this experience or its neural correlates that triggered emotional responses in law-like ways; but if Wagner experienced the gestalt of the unique unprecedented whole of that passage of music, Wagner’s grasp of that gestalt could not engage with constraining laws, because of its uniqueness and particularity.  So we must say either that Wagner’s grasp of the gestalt (as distinct from general features of the experience and of associated neural processes) was not taken into account by him and did not influence his judgment about the aesthetic merits of the passage and his consequent adoption and retention of it in his composition, or that it was taken into account, and did influence his judgment, otherwise than by engagement with general constraining laws.

          The point can be further illustrated by reference to the Game of Life.  I have already noted that gliders bring about changes only through their properties at the one-cell-and-eight-neighbours scale.  The overall pattern of a glider is an initial condition on which the rules operate, but the operation of the rules is not dependent on the overall pattern:  there is nothing in the Game that ‘takes account’ of the pattern in determining how the Game should develop.  And while, as we saw, the rules could be reformulated so as to appear to apply to gliders, this appearance was deceptive; and in any event even this appearance could not be given in relation to patterns that were rarer and more complex.  For any pattern in the Game, no matter how significant it is for an outside observer, or how striking its effects are, to the Game itself it is only an initial condition which is operated on by rules at the one-cell-and-eight-neighbours scale.  There is no chunking of the pattern or grasping of the pattern, in such a way that the pattern as a whole, or some system’s grasping of the pattern as a whole, can influence the way the Game develops.

          The question thus arises, are the originator’s or appraiser’s experiences of gestalts of Les Demoiselles or of the opening bars of Tristan merely initial conditions of this kind?  Maybe they are – although to me it seems improbable, because it seems of the essence of aesthetic judgments that they are based at least in part on gestalt experiences of the artistic work:  I will elaborate on this in Section 7.  And if such a gestalt experience is not merely an initial condition of this kind – if the gestalt has a role in that it is taken into account and thereby influences the originator’s or appraiser’s aesthetic judgment – then that role cannot be one in which the outcome is wholly determined by general constraining laws, because the gestalt cannot engage with those laws.


5   Selection:  the Third Trick Hypothesis

I think we can take it that what happens in the world is constrained by initial conditions and constraining laws, at least to the extent that there is a limited range of possible lines of development from any initial conditions, and a limited range of possible outcomes at later times:  so much is suggested by fundamental physics, including quantum mechanics.  This means I believe that the general features of any states of affairs or processes engage with constraining laws so as to limit what can happen.  If special features of any particular states of affairs or processes are also to play a role in limiting or constraining developments and outcomes, given the initial conditions, this must be by way of some kind of selection from within the possibilities left open by the general features and the general constraining laws – a selection that is not itself determined by general laws.

          It is conceivable that this selection could itself be pre-determined by the special features of the initial conditions combined with some pre-existing non-law-like constraint that engages with the special features; but if so, this constraint would not be a constraining law and would seemingly have to be arbitrary and incomprehensible.  That does not mean that such constraints cannot exist; but Occam’s razor would suggest that alternative possibilities be investigated.

If the special features of a particular gestalt experience are to be taken into account and affect an aesthetic judgment in a way that is not arbitrary and incomprehensible, then I suggest that this must be by way of a selection made by the system that experiences the gestalt, on the basis of its grasp of the gestalt.  Underlying everything that consciousness does is a vast amount of non-conscious information-processing that proceeds in a way constrained by the operation of laws on general features; and I suggest that thereby, where consciousness is involved, these unconscious processes give rise to spectra of possibilities and associated reasons for selection.  The first two tricks of consciousness bring into play the particularity of persons and experiences, so that selections can be made from these possibilities on the basis of particular grasping of particular experiences.  This I suggest is the third trick of consciousness.

           There must be a multitude of general and repeatable patterns that are, through evolution-selected processes, taken by artists and others to have aesthetic merit, and any particular gestalt will instantiate variants of some of those patterns.  An appraisal of an artistic object in its full particularity will involve some conscious or unconscious recognition of the way in which the gestalts of the work express those general and repeatable patterns, but will I suggest also involve judgments that take into account special features, including unanalysable similarities and differences between those gestalts and other expressions of the general patterns.  The general patterns would, on this approach, engage with general laws so as to constrain the appreciative response to a spectrum of possible responses, and the judgments, taking into account special features, would determine which appreciative response within that spectrum would actually occur.

          I accept that we have been programmed by evolution to find certain general configurations beautiful, and one aspect of artistic creativity must be to produce patterns incorporating general configurations that in effect push evolution-selected buttons shared by many people.  But I suggest that this may not be the whole story.  Artistic works are not mere generalities, but rather are particular instantiations, expressing some generalities in a particular way.  The relationship between the generalities and the particular expression of them can only be assessed by judgments; and I suggest that these judgments, involving as they do appraisal of a relationship between generalities with which rules can engage and particularities with which rules cannot engage, cannot themselves be constrained by general rules.  It is in this area that the artist’s or audience’s experience and grasp of gestalts plays its role.

          Judgments of aesthetic merit, particularly of innovative works like Tristan and Les Demoiselles, do not seem to be the result of the operation of rules on their general features so as to prescribe an outcome.  Such judgments may be the result of evolution-selected algorithms acting non-consciously, and indeed this is no doubt part of the story; but there are good reasons to think it is not the whole story.  If our ability to assess aesthetic merit is partly explained by our ability to grasp particular whole experiences, then our judgments about this would have to be affected by factors other than the operation of rules on general features of the object or experience in question.  That is, we must be able to take into account particular gestalts, to which no rules can apply, and assess them on the basis of qualities such as expressiveness, richness, symmetry, harmony, purity, clarity, vividness, and power to uplift or inspire or enlighten.  In such assessments, rules cannot determine the result, because rules can only engage with general features; yet the judgments are not arbitrary or whimsical.

I am not suggesting that recognition of a passage of music from Tristan, or of Les Demoiselles, requires the grasping of a gestalt.  Although these artistic objects are unique, the elements which constitute them, and by which they can be recognised, do have law-engaging generality:  this is illustrated by my table, in which every item apart from the bottom-right item has that kind of generality.  And our memory of such objects likewise can be generated through rule-governed processes, in which elements are contributed and combined in rule-governed ways.4


6  Other Judgments and Actions

So, I have argued that conscious judgments on aesthetic matters are made (1) between alternatives thrown up by non-conscious processes the unfolding of which is determined by laws engaging with general features of the world, and (2) on the basis of the conscious grasping by particular subjects of whole particular experiences that cannot engage with constraining laws.  I think this is highly plausible, for reasons I elaborate in Section 7.  And if this is the case, then it provides an example of a choice which is not wholly determined by initial conditions and constraining laws yet is non-random and rational.  And it then becomes reasonable to believe that the same is also true of other judgments and decisions that we make - such as theoretical decisions as to what to believe, particularly when there are arguments and evidence both ways; moral judgments as to what action would be right in particular circumstances; and practical judgments as to what to do in particular circumstances.  Like aesthetic judgments, each type of judgment has a normative aspect:  what is to be believed, what is to be done.  And each relates to particular circumstances, not just to general features of those circumstances.


6.1  What is to be believed

The first thing to note is that it has never been made plausible to think that justification of belief can be a matter of the operation of rules on accepted data.  Formal reasoning is only a small part of human rationality, and attempts to explain or justify our informal reasoning in terms of formal reasoning or algorithms of any kind have failed.

The most plausible suggestion as to how this might be done is that, although our informal reasoning cannot be explained in terms of rules of logic or probability or mathematics or any other rules of reasoning that human beings have formulated or discovered, nevertheless it must consist of algorithmic processes selected by millions of years of evolution.  This is no doubt part of the explanation of our rationality, but if it is put as the whole explanation, it faces two major difficulties.  First, that it does not sit well with the apparent universality of our rationality, both as between people who think in very different ways, and also in its application to problems far removed from those that faced our evolutionary ancestors; and second, that any justification of our confidence in our informal rationality must, on this approach, depend on our recognition of its having passed evolutionary tests, yet that justification itself depends on much prior application of the informal rationality it is supposed to support.  I believe a more likely possibility, which would explain why consciousness is engaged for important decisions, is that, by grasping gestalts, we can make judgments based upon this grasp, so that we can reason in ways that can’t be captured by general rules.  This, I contend, is the ultimate explanation of why, despite the arguments of Hume, Popper, Goodman and others, some conclusions of inductive arguments can be reasonable and others not reasonable (see Hodgson 1991 Ch 5, 1995, 1999).

If rationality is partly explained by our ability to grasp particular whole experiences, then our beliefs may be justified by many factors going beyond consistency with other data according to rules of logic, probability, and mathematics.  Properties of experiences such as immediacy, clarity, vividness and consistency can support belief as to the reality of the apparent objects of experience.  Beliefs can also be supported by properties such as coherence, profuseness of support, and proximity and similarity to accepted beliefs.  We can reason by analogy on the basis of unanalysable similarities and differences between some gestalts and others.  Theories underlying beliefs can be supported by properties such as coherence (again), simplicity and beauty.

          Justification of beliefs by informal reasoning will generally require attentive grasp of wholes and ‘gut feelings’ based on properties such as those I have mentioned, coupled with rule-based testing and checking.  The rule-based testing and checking can justify rejection of general beliefs, but can rarely on its own justify their acceptance.  The other procedures appear to be rational and non-random, yet not formalisable; and the fact that rules do not dictate the result does not suggest arbitrariness or whimsicality.  My own view is that this is because general principles are being applied, albeit principles that are non-conclusive and merely guide a conscious subject in making a selection among particulars, rather than determining a result on the basis of general law-engaging properties (see Hodgson 2001).


6.2  Rightness

It is sometimes said that if any meaning can be given to notions of rightness and wrongness of actions, it must be in terms of conformity to moral rules that prescribe what is to be done and not to be done.

          However, many particular circumstances occur in which the moral rules with which we are familiar conflict and thus cannot be applied mechanistically.  For example, a moral rule requiring one to fulfil commitments might in particular circumstances conflict with a moral rule requiring one to do no harm.  In circumstances like these, it appears that what is required is that one grasp the whole of the situation, in its full particularity, and make a judgment on the basis of moral principles that do not prescribe a result but can only give guidance – moral principles like do no harm, be fair, be honest, fulfil commitments, do good, and improve oneself.  So again, this suggests a role in rational decision-making for the grasp of particular wholes that do not have the generality required for engagement with general rules.


6.3  Free will

If I am right in asserting that judgments of the kinds I have discussed are rational and non-random, yet not determined by the engagement of rules, even rules of evolution-selected computational processes, with initial conditions, it seems clear that sense can be made of free will.  The outcomes of such judgments are not predetermined by anything, not even the character of the person making the judgment, inter alia because both the person and the circumstances in which the judgment is made are not mere concatenations of generalities and quantities with which rules can engage, but are particular combinations of qualities, with the particular grasp by the particular person of the particular experience of the particular circumstance making a contribution to the outcome in a way that cannot be governed by rules.

          In so far as our actions depend on our beliefs as to circumstances in which we find ourselves and as to the rightness or wrongness of actions in those circumstances, our actions depend on judgments as to what is to be believed.  And in those circumstances, it is reasonable to think also that our decisions as to whether or not to do actions, including those that we judge in this way to be right or wrong, are also non-random yet not determined by the application of laws to initial conditions.  I have previously referred to pre-conscious and non-conscious processes underlying sensual experience, recognition, interpretation, emotional feelings, and conscious deliberation. I suggest these processes produce templates for decision and action which, where consciousness is present, provide for alternatives and consciously feelable pros and cons – if only the alternatives of doing something or not doing it, or alternative ways of shaping one’s actions.  The conscious process then selects between these alternatives, whether by way of doing or not doing something, or shaping an action, or deciding upon an answer to a consciously-addressed problem.

None of this is intended to downplay the enormous roles that genes and environment and non-conscious processes have in our beliefs and actions:  as I have made clear, these factors limit the alternatives open to us, ground the reasons on the basis of which we select between those alternatives, and determine the intensity with which they are felt.  It is the selection itself, dependent on a particular subject’s grasp of a particular experience, that introduces the element of indeterministic rational choice that is the essence of free will.

          It is sometimes suggested that consciousness comes too late to control our actions, and experiments by Benjamin Libet (1978 and 1983) are cited as showing this.  One way of looking at the Libet data is to think of the non-conscious processes as ‘priming’ consciousness, rather as a pump is primed.  Because of the amount of non-conscious processing required to give rise to consciousness, there is a considerable time delay between the initial arrival of a novel sensory stimulus to the brain and its experience in consciousness:  Libet has measured this delay as about half a second.  However, if something is already being consciously experienced, there seems to be little or no appreciable delay between the arrival of a sensory stimulus appropriate to adjust the ongoing experience and the actual conscious adjustment of that experience; and this can explain why consciousness can be efficacious in the shaping of actions, as in artistic and sporting performances.  Performers emphasise the importance of concentration, and it seems clear that a performance by a pianist, for example, can be moving to the pianist and the audience because the pianist is responding to heard sounds and felt emotions in shaping the performance.

          And when acts are initiated rather than shaped, the Libet experiments suggest there are pre-conscious neural preparations of something like one quarter of a second.  On the ‘priming’ approach, it makes sense that preparation of this kind should be required before a person has immediately available the alternatives of doing something or not doing it; and if, as in the Libet experiments that demonstrated this delay, a person has adopted a readiness to do the action, it is not surprising that the available veto is not exercised.


7  Why Selection

It can be argued that to suggest that judgments are made on the basis of gestalts, in ways that are not entirely based on the operation of general constraining laws, introduces unacceptable mystery into the process.  There is some force in that argument – but I think it is outweighed by the consideration that, unless gestalts do enter into the process by which we make aesthetic judgments, and indeed decisions of all kinds, our experience of gestalts seems pointless.  The very fact that consciousness performs its first two tricks makes the third trick probable:  otherwise there would be no point to the first two.

There are two main reasons why I say that the grasping of gestalts contributes to aesthetic and other judgments.

In the first place, it seems to do so.  Of course, many things in our mental processes that seem to be the case turn out not to be, and it may be so here; but I suggest that the onus of proof lies on those who assert this.  I certainly seem to experience artistic objects and gestalts associated with them in their full particularity and uniqueness, and I seem to attend to this experience in whatever appraisal I make of them.  I certainly do not consciously apprehend such objects as amalgams of types of patterns and associations; although my unconscious processes must to some extent at least proceed on this basis.  Rather, I apprehend them as unique particular objects, and have a clear impression of them as such; and I cannot conceive of them affecting me as they do without my apprehending them in this way.  Unique particular wholes is what they are to me, and it is as such wholes that I enjoy them and appraise them.

I cannot believe that a unique work like Tristan or Les Demoiselles is created without the artist experiencing it in its uniqueness and particularity, and assenting to and adopting it on the basis of that experience.  And although I have been focusing on quite complex objects, in order to make it clear that there can be no law of nature which applies to them as wholes or gestalts, the same argument also applies to much simpler objects, such as a simple melody.  I believe it would be bizarre to suppose that a melody like the Air from County Derry (‘Danny Boy’) has, as a gestalt, generality such that general constraining laws could engage with it or with gestalt experiences of it or of substantial parts of it, as opposed to engaging with general patterns instantiated in it.  And I believe it would also be bizarre to suppose that, in responding to this melody, we are not, to some extent at least, responding to our experience of gestalts of the melody and/or substantial parts of it.

A second reason for thinking that we do respond to and take into account particular gestalts is that this helps to make sense of why we have consciousness, and what it is for.  To the extent that objects and processes (including conscious processes) have law-engaging generality, to that extent the contribution they make to the way the world unfolds could be contributed without consciousness:  the neural correlates of any conscious objects and processes could make precisely the same contribution that conscious objects and processes make.5

          However, to the extent that the first two tricks of consciousness give rise to an experience of a gestalt that does not have law-engaging generality, such an experience could make a contribution to the way the world unfolds that could not be made by its neural correlates, as the system that has the benefit of the unique chunked experience responds to it in its full particularity.  The particular unified global gestalt is experienced by a particular subject, 6 who can respond to it in its full particularity.

          It makes sense that consciousness should have a special role of this kind.  Non-conscious information-processing, as we understand it, works perfectly well with numbers and symbols and codes, and our computers translate these things into natural languages and images only for the benefit of human operators.  Our brains do process vast amounts of information unconsciously, presumably in computer-like ways, and also produce much bodily activity as a result.  But our brains also go to enormous trouble to produce conscious experiences, which present vast amounts of information most relevant to choices that need to be made, in simple and striking ways.  They also seem to be structured in such a way that, when a problem arises that needs our best response, our conscious attention is immediately brought to bear.

          I believe the explanation for this is that, just as our aesthetic judgments benefit from the grasp of unique gestalts, so also do our prudential judgments and actions.  For this reason, the three tricks of consciousness have been useful to our evolutionary ancestors, and so have been selected and promoted by evolution.7


8  Concluding Thoughts

          On the approach I am suggesting, there is room for a radical kind of emergence:  not merely the occurrence of new patterns on which the same laws operate in the same way, or even which engage new laws that have never before had occasion to operate; but rather, the occurrence of patterns involving the experience of particular gestalts by particular subjects, that can then contribute to the way the world develops in a manner not entirely determined by general laws.

One could envisage this situation emerging gradually in evolution, as the two tricks of consciousness gradually develop in self-organising systems; and one can envisage it emerging in the life of individual persons.  Quantum mechanics tells us that initial conditions and general laws do not uniquely determine outcomes.  Initially, one may have systems where initial conditions and constraining laws leave open spectra of possible developments, one of which occurs at random.  Gradually, self-organisation and then conscious choice may move into this space.  I am not suggesting the intervention of an immaterial subject or soul; and there need be nothing to distinguish between systems that constitute persons apart from their objective properties, the general features of which engage with constraining laws so as to constrain outcomes to a spectrum of alternatives, and give rise to the first two tricks of consciousness.  These general features thereby make possible the third trick of consciousness, constituting a particular person with the capacity to choose, and providing the consciously-felt reasons for choice, including gestalt experiences – and the exercise of this capacity on the basis of these reasons is then a matter of judgment, not pre-determined by constraining laws or any other rules.

          My suggestion is that our actions are shaped by nature, nurture, and choice.  Nature and nurture (and to some extent previous choices) supply the alternatives and the reasons, and we resolve them in ways that are neither random nor pre-determined by initial conditions and constraining laws.  At the margins, we can choose and shape the world we want, and we can choose and shape what it is we want.  We are constituted by our general features in their unique combination, which can be defined quantitatively, but which (as wholes) have irreducible qualitative features.  The unique subject combines with unique experience so as to give rise to alternative lines of development constrained by general features and general laws, one of which is selected by the subject on the basis of the particular experience.



1.  In saying this, I am not asserting that states of affairs are basic, and objects and processes are secondary:  it is simply convenient to conduct this discussion in terms of states of affairs. 


2.  In particular, in that it leaves out of account many other law-like processes involved in the perception of a work of art, including for example those involved in recognition of types of objects and in the emotions triggered by such recognition.


3.  The point here bears some relation to the ancient problem of epiphenomenalism, but with important differences; in particular in that it leaves open whether items on the right, apart from the bottom one, could have law-governed effects.


4.  To the extent that a judgment about the aesthetic merits of an artistic object is based on memory, then, it could be argued that the original grasp of the gestalt experience is superfluous.  I would deny this, for two reasons:

(1)  The original grasp itself and the person’s response at the time make their own contributions to the appreciation process.

(2)  The processes that enable grasp of a gestalt in one’s memory are of the same kind as those that enable the grasp of the gestalt of the original experience; so in that sense, there could not be the grasp of the gestalt in one’s memory unless there is a process enabling this grasp that could also have enabled the grasp of the gestalt in the original experience.


5.  This is similar to a point made by David Chalmers (1997) at 27-28.


6.  I’m not suggesting that the subject is distinct from the system in which the chunking occurs or that it is unchanging, much less that it is an immortal soul; although my approach does suggest that there must be a continuity of the subject as it experiences a gestalt and responds to it.


7.  Much the same is true of other animals that are conscious, but I do not attribute free will to them as to us, because we are different in the richness of the alternatives and supporting considerations, and the depth of our grasp and understanding of them.



Chalmers, D. (1997), ‘Moving forward on the problem of consciousness’, Journal of Consciousness Studies, 4 (1), pp. 3-46.

Davidson, D. (1970), ‘Mental events’, in Foster, L. and Swanson, J. (eds.), Experience and Theory (Amherst:  University of Massachusetts Press).

Hodgson, D. (1991), The Mind Matters (Oxford:  Oxford University Press).

Hodgson, D. (1995), ‘Probability:  the logic of the law - a response’, Oxford Journal of Legal Studies, 14, pp. 51-68.

Hodgson, D. (1999), ‘Hume’s mistake’, in Libet, B., Freeman, A., and Sutherland, K. (eds), The Volitional Brain (Thorverton:  Imprint Academic).

Hodgson, D. (2001), ‘Constraint, empowerment, and guidance:  a conjectural classification of constraining laws’, Philosophy, 76, pp. 341-70.

Libet, B. (1978), ‘Neuronal vs subjective timing for a conscious sensory experience’, in Buser and Rougel-Buser (eds), Cerebral Correlates of Conscious Experience (North-Holland, Amsterdam).

Libet, B., Gleason, Wright and Pearl, (1983), ‘Time of conscious intention to act in relation to onset of cerebral activities (readiness potential):  the unconscious initiation of a freely voluntary act’, Brain, 106, pp. 623-42.

Levy, S. (1993), Artificial Life (Harmondsworth:  Penguin).

Poundstone, W. (1987), The Recursive Universe (Oxford: Oxford University Press).


Return to Home