The sleep and the dream of reason

1. Introduction

In 1799, Goya published a set of 80 aquatints under the title Los caprichos. Number 43 was called El sueño de la razón produce monstruos (The sleep of reason produces monsters). It may be found in many places on the web, for example:

https://eeweems.com/goya/sleep_of_reason.html

As always with worthwhile art, the picture can be read in several ways. In this post we shall explore a few of the thoughts we might have, over 200 years after the picture's creation.

Our focus will be on what may be directly relevant to our current world. We shall therefore leave aside the acerbic comment on the state of politics and society in Goya's own time that is made by Los caprichos as a whole. We do so not because we have no such problems today, but because the problems and the targets of criticism would differ sufficiently to require different lines of attack. More generally we shall not try to deduce Goya's own intentions, nor elaborate on the artistic context. We shall cheerfully take the single picture that interests us out of context. And in our exploration of its messages, we shall reduce those messages to prosaic formulations. We must do so to discuss them, but in doing so we shall inevitably lose the artistic richness of the picture. Indeed, what we say will be based on the picture's title and some accompanying text. We shall not explore the visual detail of the picture itself, but only note that it makes the content of the title and the accompanying text vivid.

2. The title and the text

There is text to accompany the picture. There are some different versions in circulation. Here is one version:

La fantasía abandonada de la razón produce monstruos imposibles: unida con ella es madre de las artes y origen de sus marabillas.

Imagination abandoned by reason produces impossible monsters: united with her, she is the mother of the arts and the origin of their marvels.

We shall allow this text to guide our thoughts. In particular, we shall use it to separate two different threads that are indicated by two meanings of the word "sueño". It can mean either sleep or dream.

If "sueño" is read as "sleep", the message of the picture's title (but not the only message of the picture itself) would be that it was dangerous to leave imagination to its own devices with no rational control. That reading will lead us to the thread of Enlightenment, which we shall follow in section 3.

If "sueño" is read as "dream", the title could well be read as saying that reason, active in its dream, would itself produce monsters. That reading will lead us to the thread of the perils of reason, which we shall follow in section 4.

If we were to be guided by the text, its reference to imagination's being abandoned by reason would rule out a reading of "sueño" as "dream", or at least require any such reading to involve the stilling, and not the activity, of reason. This would limit us to following the Enlightenment thread. But we need not be bound by the text. We therefore have scope to read  "sueño" as "dream". Specifically, the application of reason might have monstrous consequences when reasoners lacked the imagination to see the monstrosity.

3. The sleep reading

3.1 The need for reason

One reading of the text that accompanies the picture is that imagination can produce wonderful results if it works hand in hand with reason, but can lead us disastrously astray if it does not. That thought would be entirely in line with Enlightenment values, which vaunted reason but did not advocate a Puritan iconoclasm in which there would be only cold reason and no art.

The particular form of going astray that is most strongly suggested is that of being driven to unwise choices by having irrational fears or by the acceptance of superstition, things that could both stand in the way of social and intellectual progress. But there are other forms too. One might for example adopt some pseudo-scientific theory and be diverted from progress in that way. Pseudo-science is not a scary monster to the one who adopts it. It may even be very appealing, because it may offer an easy route to apparent knowledge. But it is still a monster by virtue of its standing in the way of human progress.

3.2 Which should be in charge?

The text speaks of uniting imagination and reason. It does not say whether either of them should be in charge. An Enlightenment approach would require reason to be the one in charge if either were to be. But it would also be open to a partnership in which neither would be in full control, imagination would bestir a reason too dull, and reason would discipline an imagination too wild.

There would still be asymmetry. Bestirring is not the same as disciplining, and only the latter is straightforward to see as a form of control. It is about restraint. In addition, we can see it as amounting to the giving of specific instructions about how far to go in imagination and which topics not to stir up with fantasies (for example because there is no scope for reasonable doubt about conclusions already reached). Bestirring, on the other hand, is about breaking restraints. It is also non-specific as to what may emerge from reason once it has digested some new fantasy. Having said that, to bestir is to reject complacency. In that respect, to bestir is to tell reason in general terms how to work. And in section 4 we shall give imagination a role in guiding reason away from certain avenues that might tempt it.

In order for a partnership of reason and imagination to be satisfactory, it would be important for reason not be be too controlling. Given that the objectives of an Enlightenment mindset include not only the elimination of superstition but also human advancement, imagination must be given scope to push thought forward in unpredicted ways. This is the only way to advance into intellectual territory that is currently unknown. Indeed, reason in a broad sense will see this and will restrain its own impulse to restrain.

Such a broad notion of reason is perfectly compatible with the Enlightenment message of Goya's title. His concern is with the danger of reason in general falling asleep. The sleep of the restraining aspect may be the source of the danger the title identifies, but that does not exclude there being other aspects which, for reasons not indicated by the title, it is important to keep awake.

3.3 How reason might exercise control

One way in which reason might exercise control would be to review the results of imagination's work, discarding some results, lauding others, and saying that yet others were worth further development to see where they might lead. Reason would sometimes make mistakes, closing down some avenues that would in fact have been productive and encouraging work on other avenues that were in fact dead ends. But it would be part of the improvement of reason to reduce the number of such mistakes.

Reason might also guide the development of promising lines in some detail, specifying which aspects needed attention and selecting between possible developments that imagination had so far only sketched.

This raises the question of what we should regard as the talents of reason. In order to fit in with the sleep reading of "sueño" and still to see reason's sleep as problematic for an Enlightenment agenda that seeks the minimisation of error, reason must have a role of criticism. But it might also have a creative role, not leaving all of that to a separate faculty of imagination. Then it could do more than merely select between possible developments. It could suggest some developments and sketch ways in which they could be carried forward, then hand the task of full development back to imagination.

This would blur the boundary between reason and imagination. That would in turn weaken the criticism of pure reason that we shall explore in section 4, the criticism that an absence of imagination would lead reason to make proposals which would have been detected as unacceptable had an imaginative appreciation of how things would be for people played a role. And the warning about uncontrolled imagination we are taking from Goya in this section would not apply to imagination generally, but only to the imagination that was outside reason. The message about the sleep of reason producing monsters would however still be appropriate, because it would apply to the part of imagination that was outside reason when reason was not awake to exercise control. And the part that was outside reason would still be the major part of imagination.

There is a good case for blurring the boundary and admitting some imaginative component to reason. We can see how things may be if we do not blur the boundary by looking at David Hume's comment that "It is not contrary to reason to prefer the destruction of the whole world to the scratching of my finger" (Hume, A Treatise of Human Nature, Book 2, part 3, section 3, paragraph 6). This remark could only be made on the basis of an implausibly narrow concept of reason. Hume had good reason to use such a narrow concept. He wanted to separate out the role of the passions, and to show that reason with no tincture of passion would get us nowhere. But we can respond that his narrow concept of reason differs from our ordinary concept, which allows us to speak of rational desires and impulses, and to call at least some of them rational because they come from within reason rather than merely because they would pass reason's scrutiny.

4. The dream reading

4.1 The need for imagination in human affairs

Careful reasoning about the management of human affairs can have repugnant consequences, starting with an array of petty controls for the public good and ending with the horrors of communism. In between we have things like bringing up people to be organ donors because the benefits to the organ recipients will be very great, as in Kazuo Ishiguro's novel Never Let Me Go and the film made of it. And at the end of the range that is superficially acceptable but on reflection ethically dubious, we have Condorcet's calculation of the acceptable probability of convicting innocent persons, a calculation that has been characterised as showing "everything that is at once attractive and repellent" about the Enlightenment (Daston, "Condorcet and the Meaning of Enlightenment", page 113).

4.2 How things would be for people

4.2.1 The need for a form of empathy

One aspect of the problem of unbridled reason is that reason might calculate away, with no reality check that was grounded in an appreciation of how certain actions or events would be for individual human beings. Such an appreciation would be a form of empathy. And it would not be adequately characterised as an appreciation of someone's level of happiness such as a scientific utilitarian would like to measure. That would not be enough because how something would be for a person would not be a single measurable feature of his or her mind, and perhaps not even a function of several such features. Maybe it would be emergent from measurable features of the mind, and maybe it would be independent of such features. But in any case, it would not be something that the scientific manager of people or society could grasp without having the empathy that would come from reflection on the experience of human life from the inside. One would need an imaginative leap. It would not need to be fully into the heads of people, but it would have to go as far as grasping how things would be for people if certain changes were made.

Here we may locate one need for imagination to steer reason. In so doing, we read "sueño" as "dream". The rational and enlightened social reformer may be in a dream that reason suffices to give us all the answers. The reformer may be cut off from the human aspect of reality by the fact that the empathy that requires imagination is left outside the narrow compass of the dream, just as many of the practical constraints of reality are left out of actual human dreams.

4.2.2 How imagination might work

We are here asking imagination to draw attention to how things would be for people in general, or perhaps for people with the same cultural backgrounds as those who would be affected by a proposal. (In a mobile world it would be backgrounds in the plural, and one should also consider people from additional backgrounds who might come to be affected in the future.)

Here a problem arises. Imagination would need to range very widely even to identify candidate pictures of how things would be for people of all relevant cultural backgrounds. And a long list of candidate pictures would not suffice. For imagination to keep reason grounded in a useful way, it would have to identify how things would actually be for people and not random ways in which things might be for them. That would be difficult. Even people with the same cultural background as oneself can differ enough for it to be hard to be confident of how something would be for them. But a long list of random ways in which things might be for other people would mean that almost every proposal would be identified as possibly problematic for someone. Then nothing would ever get done.

Fortunately, a goodly proportion of the benefit of imaginative control over coldly calculating reason might be obtained more easily. Simply asking how things would be for oneself would be a way of detecting some dangerous proposals. This would not however be enough if one had an odd personality, or if the proposal would be of benefit to oneself and of detriment to others. Then it would be vital to consider how things would be for other people. But much of the benefit might still be obtained from consideration of how things would be for people with one's own cultural background. This would not always be enough. Some proposals would play out very differently in different cultures. But it would sometimes be enough, because while our cultures differ, we have a great deal in common simply by virtue of being human.

The significance of human commonality can be brought out in another way, arguing it to be necessary rather than sufficient. Martians, with a complete grasp of our neurons but no experience of human life from the inside, would not be able to use imagination in the way we require to keep calculating reason in check. This is a bold claim to make when one is not oneself a Martian, but it is a plausible one.

Although we ask imagination to draw reason's attention to important considerations, and it would have to do so by formulating concerns explicitly, we have also proposed that how things would be for people is not to be computed from measurable characteristics of their minds. If that is so, an appreciation of how things would be is not to be derived solely from a theory of other people's minds. To borrow from the philosophy of mind, a form of simulationism rather than theory theory would be the appropriate conceptualisation of what went on when someone's imagination was discovering how things would be for people. It would however be simulation of feelings rather than of deciding how to act. And there is the proviso that simulation, whether of others or of one's own future self, would not need to do all the work. Theorizing that was grounded in simulation might do the job.

4.3 The acknowledgement of possible error

Another vital element in sensible thought that may get left out of reason's dream world is awareness that one might be mistaken. This omission may well be linked to the omission of imagination, because the thinker may fail to imagine specific ways in which he or she might be mistaken. Such specific ways convey awareness of the risk of mistake powerfully. If someone only has a non-specific general recognition that mistakes could be made, the risk may well be overlooked.

The result of a failure of awareness that one might be mistaken can be a disastrous dogmatism, with counter-arguments to the recommendations reached by narrow reason being dismissed simply because it is thought that the recommendations cannot be mistaken. Dismissal may be effected by using inadequate arguments which are thought to be adequate, or simply by ignoring the counter-arguments.

A rational faculty of the highest quality would not fall into this trap. It would be clear that error was always possible and that dogmatism was therefore inappropriate. But our intellectual capacities are limited. We may be so busy making sense of all the complex data that the world offers us that we fail to explore the scope for error, or even forget it entirely.

For imagination to be effective in drawing attention to specific ways in which reason's conclusions might be mistaken, it would need to be subject to its own form of rationality. It is easy to invent random possibilities for error, but many of them would lack plausibility. Implausible possibilities would easily and rightly be dismissed by defensive reason. It is harder to concentrate on plausible possibilities. Nor would it be enough to generate a vast number of possibilities and rely on the expectation that some would be plausible. Reason would not have the patience to sift through them all.

Here, as in section 3.3, we blur the boundary between reason and imagination. But here it is for a different reason, and the blurring is to build a bit of reason, the bit that identifies plausible possibilities for error, into imagination, rather than as in section 3.3 build a bit of imagination into reason.

4.4 Arts and sciences

So far in this section we have been concerned with the need for imagination in order to provide a reality check on rational social reform. Now we shall turn our attention to the danger of reason's dreaming that it suffices in the arts and the sciences.

There is little danger of reason's dreaming that everything is solved, and that no further progress is necessary. There is greater danger that reason will become set in its ways and will believe that these are the only useful ways. Work in the natural sciences must be conducted like this. In the social sciences and the humanities, the following concepts should be deployed. In the production of art and music, these are the norms of composition to be followed. If reason dreams that it knows how people must work, it may miss out on opportunities to make significant progress. It is imagination that can shake reason out of such a complacency of approach.

The danger is present in all areas of work. But it is not great where it is of the essence of the area of work to define problems and then solve them. The natural sciences, and to a lesser extent the social sciences and the humanities, are like this. The insistence that solutions be found creates its own pressure to consider new ways of working. Imagination is needed to devise such ways, but reason of any sort that would be satisfactory for these areas of work could not get itself into so deep a dream as to ignore the need to consider new ways of working.

The danger of not introducing new ways of working is greater in the creative arts. To take an example from painting, it could have stayed representational. Then it would have missed out both on quasi-representational approaches such as Cubism and Expressionism and on the purely abstract. Its advance would have been sharply slowed, even though it could for example still have progressed into Impressionism, a way of working which in at least some of its forms continued to give priority to accurate, although not precise, representation.

Without occasional doses of inspiration to work in new ways, reason that dreamt it had the right approaches all worked out would move the arts slowly towards producing not living monsters, but fossils. Then the arts might become monstrous caricatures of the glorious enterprises they once had been.

We may also make a connection between the role of imagination in the creative arts and the role of imagination in keeping the work of reason in managing human affairs from straying into monstrous proposals. There, imagination was needed to give a role to an appreciation of how things would be for people. In the creative arts, quality is not correctness in any scientific sense. Rather, it is an ability to speak to people, to bring out and develop what they see in themselves and in others. A non-human artist, one who did not know human life from the inside, would be unlikely to produce much work that would speak to us, although they might occasionally have the good fortune to do so. A human artist who can imagine the responses of people to his or her work has a far better prospect of success. We may add that how something might strike the artist himself or herself is likely to be an inadequate guide. An artist who knows what messages he or she intended to convey will be more likely than others to see those messages in his or her own works. He or she may then fail to see that some of those works will not speak to others.

5. In praise of imagination

One theme we have not discussed so far is the intellectual glory that imagination may allow us to achieve. The text accompanying the picture tells us that imagination united with reason "is the mother of the arts and the origin of their marvels".

This is plainly true. And sometimes reason should fall asleep and give imagination free rein, before of course waking up and reviewing the results carefully.

Here is a short poem on Goya's picture that I wrote several years ago.

The sleep of reason tempts away from thought
It opens up the doors to let perceive
Such spectres new, in net of fancy caught
That we, in sensing, take of senses leave.
The reason for this sleep we thus behold
It lifts us up from bed to heaven's light
If we knew only that which reason told
Our vision would be empty as the night.

6. References

Daston, Lorraine. "Condorcet and the Meaning of Enlightenment". Proceedings of the British Academy, volume 151, 2007, pages 113-134.

Goya, Francisco de. El sueño de la razón produce monstruos. https://eeweems.com/goya/sleep_of_reason.html

Hume, David. A Treatise of Human Nature. Available in several editions.

Ishiguro, Kazuo. Never Let Me Go. London, Faber and Faber, 2005. (A film directed by Mark Romanek was released in 2010.)