EMOTIONAL TRUTH

 

to appear in an Aristotelian Society symposium (joint session) with Ronald de Sousa

 

 

II – Adam Morton

 

EMOTIONAL ACCURACY

 

ABSTRACT   It is accuracy rather than truth itself that is valuable.  Emotional truth is a dubious though attractive notion, but emotional accuracy is much easier to make sense of.  My approach to accuracy goes via an account of what makes a story accurate.  Stories can be accurate but not true, and emotions can be accurate whether or not they are true.  The capacity for emotional accuracy, for emotions that fit a person’s situation, is an aspect of emotional intelligence, which is as important an aspect of rational human agency as the intelligent formation of beliefs and desires.

 

I

 

Cheap truth.   Truth comes in many forms, some cheap and some valuable.  Distinguish two dimensions of cheapness.  One dimension extends in the direction of vagueness, indefiniteness and generality.  If I claim that some flowers are coloured, or that music is sometimes nice, what I say is true, but cheaply so.  Another dimension extends in the direction of the range of attitudes that can be counted as true.  Truth can be extended from assertive sentences to beliefs to questions and requests at very little price.  When a person attitudes that p, and p, we can count her attitude as true.  So a Yes-No question is true if the answer is yes; a desire is true when it is satisfied.  And we can say that all Jane's desires on Tuesday were satisfied, which would be equivalent to ‘if on Tuesday Jane wanted cats to fly then cats flew, and if she wanted 34+76 to be 994 then 34+76=994, and ....’.  Similarly, we can say that ‘Hamlet killed Polonius’ is true iff Hamlet killed Polonius, and that ‘eπi   = -1’ is true iff eπi   = -1, without worrying about where in the world to find Hamlet, Polonius, and imaginary numbers.  None of this is very demanding; the conceptual price is low, as the minimalist literature shows.  (See Williams 1976,.Horwich 1998.) 

            Emotional truth is easily achieved if one wanders far enough out along these dimensions.  My fear that the dog will bite me is true if and only if the dog will bite me.  My elation that life has many joys and my depression that life is a grim business are both true since life is a grim business with many joys.  But there’s no philosophical pride to be had from bringing home these trophies; any child with a butterfly net could have gone out and got them. 

            Now to the more valuable kinds.  The opposite of vagueness is precision, and precision combined with truth gives accuracy.  Accuracy certainly adds value to truth.  For one thing it allows non-perverse speculation: the difference between scientific cosmology and metaphysical rambling is that cosmology distinguishes between finely differentiated hypotheses - whether fundamental constants have this value or this slightly different one - and tries to distinguish the different consequences they would have.  And on the other dimension, the opposite of minimalist content-matching is to insist on a world-to-mind direction of fit in which determinate aspects of the state have to match determinate aspects of the world.  (A substantive theory of truth – correspondence, as I’m slanting it – thus aims not at telling us what propositions are to count as true, but what kinds of truth to count as valuable, a point ignored by Lewis 2001.)

            Emotional truth that had these value-adding features would be something to aim for.  There would be a point to directing the evolution of our emotional states towards it, just as there is a point to directing the evolution of our beliefs towards the more valuable, but only the more valuable, forms of truth.  Analogous to the way precision in theory allows responsible speculation, precision in emotion allows responsible intensity.  If you have the exact emotion for the situation, then you can feel it whole-heartedly, without the danger of inappropriate blundering.  A bull that dances through the china shop.  And analogous to the world-to-mind fit of beliefs would be some notion of an emotion that is demanded by the situation.  Elation where elation is right, depression or anger where that is right, whether or not the person has grounds to motivate their feeling this right thing.

            These remarks are meant to elicit sympathy for de Sousa’s project.  To the extent that we have a grasp of the right emotion for a situation, the objectively right emotion, we can see analogs in emotion of the valuable features of true belief.  But they are also meant to insinuate a doubt.  The intuitions are linked not to the core idea of truth itself but to the value-adding aspects that make it worth having.  In this paper I shall argue that some of these aspects are independent of the core.  We can make sense of emotional accuracy without having to make sense of emotional truth, at least not in more than the cheap and easy way just described.  Some of the consequences of accuracy-without-truth, though, are in many ways like those that de Sousa wants from emotional truth.

 

 

II

 

Accuracy without truth.   Consider two stories.

Story 1:  a carriage rolled north down Baker Street through a thick London fog on a cold December day in 1887.  As it came to Marylebone Road the passenger rapped on the driver's window and asked to be let out.  Only the most acute of observers would have recognized the crippled Crimean war veteran who emerged as the famous detective Sherlock Holmes.

Story 2:  a boat drew slowly along the Baker Street canal in the balmy weather of London in the winter of 1887.  As it joined the Thames a passenger leapt to the bank.  That person continued his journey on foot.

Neither story is true.  Possibly neither is false.  But the first is in two respects more accurate than the second.  Baker Street does not have a canal, and even if it did it would not reach the Thames.  The winter of 1887 was not balmy.  That is the first accuracy, fit: the first story fits the world as it is, even though it does not say anything true about it.  The first story is also detailed in a way that the second is not: it gives a specific name to its protagonist, and describes his appearance.  Though both stories can be matched with many non-actual worlds, the first applies to fewer than the second does: it is more restrictive.  (We are probably speaking of infinitely many worlds in both cases, so ‘fewer’ is problematic.  It would be best to consider cases where one story’s worlds are a subset of those of another.  But that would require four stories rather than the two I used.)

            The two aspects interact.  Detail allows fit.  If a story has enough details that can be taken as true of an actual situation then it will fit it.  Fit selects detail.  If a story is taken as fitting a particular situation then we can assess the detailedness of its description of that situation.  This suggests a tentative definition of accuracy.  One story is more accurate depiction than another of an actual situation when there are more elements of the one that are true descriptions of aspects of the actual situation than there are of the other.  (One story might be taken to be intrinsically more accurate than another when there is an actual situation such that there are more elements of the one that are true descriptions of aspects of that situation than there are elements of the other that are true descriptions of any actual situation.)  That will do for now; the definition is not meant to be taken very carefully.  (It surely will not survive rough handling: taking stories as closed under logical consequence and then literally counting true sentences, etc.)

            Some think that stories are true of worlds, and thus simply true when they are true of the actual world.  I do not want to get into this question.  The important point is that one not-true story can be more accurate than another.  Science fiction is not very accurate, at any rate not accurate about the technological possibilities (or even usually the laws of nature) of the present actual world.  Cowboy fiction is said to give a very inaccurate impression of life in the Wild West.  Zola or Hardy probably do give relatively accurate reports of life in the times and places they discuss.  But none of these stories are true.  In fact, a story can have a good measure of accuracy while being not just true but impossible.  Kurt Vonnegut’s Cat’s Cradle is full of historical, sociological, and emotional accuracy while describing something that just can’t happen.

            Accuracy as just described seems to presuppose truth.  An accurate story has many elements that are true descriptions of an actual situation.  But a more careful formulation takes care of this.  A story can be taken as describing a situation no elements of which does it actually name.  For example a story might begin ‘The general had accumulated many powers, so many that concerned citizens plotted to assassinate him’.  It might be taken as describing events in Rome in the first century BCE, or in many other times and places.  But no element of it is simply true.  Conversely a Jonathan Miller type production of ‘Julius Caesar’ might add enough detail that – incorporating all elements of the production into the story – it was an accurate portrayal of Tony Blair and his entourage.  The assassination itself would then be a non-descriptive detail that gained significance from its links to the descriptively accurate elements.  Neither accuracy nor truth simply presupposes the other.

 

II

 

Emotional accuracy.   What does this have to do with emotions?  The essential link is that a person’s emotions involve representations, and these representations can be more or less accurate depictions of her situation.  Contrast two classes of cases. 

(1a) An engineer is laid off by her company.  She realizes that the economic climate is not good for getting another job of the same kind, feels relieved that she does not have to face more boring programming disguised as design, and goes back to university to do a MBA.

(1b) An engineer is laid off by her company.  She takes this as showing that she has neither the technical nor the personal skills for success in a demanding profession, becomes very unhappy, and does not look for another job.

(1c)  An engineer is laid off by her company.  She reflects on the less competent and less hard working colleagues who have kept their jobs and of the lack of respect her boss has always shown to her.  She gets very angry, goes into his office and pours a cup of coffee over his head.

 

(2a) An engineer is laid off by her company.  She thinks of all the desired things that will now never happen and is overcome with sorrow.  She becomes very unhappy at the fate of abandoned animals, and cries whenever she sees a dog walking without a leash, or a non-fat cat.

(2b) An engineer is laid off by her company.  The next day she finds her mind is full of confusing thoughts moving in all directions.  There is something exciting about the confusion and she develops an enthusiasm for the company.  She starts a web site on which satisfied customers and grateful employees can register their good feelings.

(2c) An engineer is laid off by her company.  Feelings of anger rise up in her and she directs them at American policy in the middle east.  She becomes a fervent campaigner for the internationalization of Jerusalem.

 

            The cases under 1 resemble story 1 in a way that the cases under 2 resemble story 2.  That will only be true of some ways of imagining the cases, filling in the details.  But they are the natural ones, the ones that would first occur to one.  Suppose that we have a detailed filling in of one of these cases, including on the one hand the engineer’s beliefs, intentions, fears, and desires, and on the other hand her whole physical state and the state of her environment, the sensations she experiences, and the basic acts she performs.  Call the first of these ‘the emotion-story’, and the second ‘the situation’.  Then the emotion-stories of the (1) cases are more accurate depictions of the situation than the emotion-stories of the (2) cases.  More of the facts are accurately represented in these stories.  Consider (1a).  The engineer's emotion is one of relief and re-directed interest.  These emotions are directed at specific aspects of her situation and do not make sense without them.  They involve (or require, or even consist in) beliefs about the character of her work before she was laid off, beliefs about the character it would have assumed had she been one of those not laid off, desires to do one kind of work rather than another, intentions to act in one way rather than another, and so on.  Contrast this with (2a).    The engineer's emotion is one of sorrow directed at the plight of animals.  But, at least on one natural way of filling in the details, there are no specific episodes of animal suffering that give detail and specificity to the emotion: many associated beliefs are not true of the engineer’s life, and many associated desires do not lead to successful acts.  (They're not true desires, in the cheap way of speaking I suggested above.)  This is generally true of natural ways of imagining the (2) cases: they do not latch onto actual features of the situation as it is.  In fact, in order to imagine oneself into the situation of the engineer in the (2) cases one has to imagine her misconstruing and misrepresenting what is going on and what the connections between events are.  This is much less so in the case of the (1) cases.  The emotions there not only are sustainable in the face of an accurate grasp of the facts and possibilities, they build on a network of representations of the details of the person’s situation.

            I am trying not to put this in an overly cognitive way.  On a 1970's-type account the emotion just is a complex of states essential members of which are propositional attitudes, which in accurate cases have true propositions as their objects.  I take it that a number of writers, notably de Sousa and Greenspan (see de Sousa 1987, Greenspan 1988, also Goldie 2000) have shown us more plausible ways of recognizing that thinking is essential to emotion without turning emotions into thoughts.  Without taking on the details of any of these accounts I shall assume that when one is in an emotional state there are patterns of belief and belief change, desire and desire change, and characteristic intentions, that are essential to ones being in that state rather than another.  If a person is afraid then there is a pressure towards thinking that some things are dangerous, and a tendency towards wanting to avoid or escape some things, whether or not she succumbs to the pressure or goes along with the tendency.  This is enough to make what I have called the emotion-story essential to the emotion, and thus to give the emotion an intrinsic degree of accuracy as a depiction of a person’s situation.

            I said that accurate emotions are sustainable in the face of an accurate grasp of the facts and possibilities.  Why possibilities?  Consider someone who takes as fearful something that cannot hurt him, or who greets with joy something that cannot do him any good.  The emotions don’t fit the situation not because the object will not harm or will not help; after all, it is appropriate to be afraid of a rattlesnake that in fact does not bite one.  The lack of fit comes because something is thought to be capable of what it is not.  More generally, an emotion can be inaccurate because it misrepresents the possibilities of the whole situation.  Most emotions are action-guiding, taking action in a very general way to include strategies of thought.  (This is a central idea in most of the papers in Goldie 2001.)  They will not serve this role if they are unhinged from the actual situation of the agent; and they will not serve it if they do not respect what is actually possible and impossible, in fact what possibilities are more or less remote.  So an accurate emotion must not only contain detailed representations that fit the actual situation; it must represent that actual situation as rightly situated in the galaxy of could-have-beens and would-have-ifs around it.  This might seem to distinguish accuracy of emotion from accuracy of belief.  I think it does not, though.  A belief can be inaccurate though true in a detailed way of the actual world.  Consider for example a rich and complete system of unnatural Goodmanian concepts, cutting across natural kind boundaries in weird and peculiar ways, and consider beliefs expressed in terms of them.  The belief that all emerats are granimals is true (emerats are emeralds that come to human notice before 1 Jan 3000 or otherwise rats, and granimals are green things noticed before that date or otherwise animals).  But it misrepresents what emeralds and rats are like and taken together with other similar beliefs would misrepresent what is possible for them.  So respect for how a situation is situated among its possible variants is something we should write into a better definition of the accuracy of belief, too, taking accuracy even further from truth.

            Accurate emotions are not well described as true.  After all, the analogy is with an accurate story, and many very accurate stories are not true.  The difference shows up in the non-uniqueness of accuracy.  All of (1a)-(1c) are accurate, accurate to the same facts about the engineer's life.  I see no reason to think that any one has to be more accurate than the others; each could invoke as rich a body of beliefs and desires, fitting the person’s situation and its possibilities as well as each other one.  (That is why (1b) is included: emotions that we think of as less wise or less admirable may still be accurate.  But see section III below.)

            Another way of putting it.  An accurate emotion is like a rich myth, deeply engaged with the details of some aspect of the world.  A less accurate emotion is like a shallow or artificial myth, a Walt Disney substitute, which tries to depict mythical events that bear no detailed relation to what actually happens in people's lives.  Or, the accuracy of an emotion is like the observational accuracy of a scientific theory, which can capture actual and potential observations more or less well.  Theoretical and observational assertions and concepts can be intimately connected; neither may be intelligible without the other, and yet it is clearly true that observational accuracy does not guarantee truth.  There can be rival equally observationally accurate theories, relative to any way of drawing the somewhat arbitrary line between observation and theory.  And among non-true theories some will be more accurate observationally than others.  For some purposes, e.g. navigation or bridge construction, observational accuracy will be more important than truth.  We want a rich and reliable body of connections with the ways the world impinges on us.  So too with emotions: among the variety of attitudes we could take to the situations we find ourselves in, we want those that give a rich and reliable set of connections to guide our further acting and feeling.

 

III

 

Emotional intelligence.  My main point has been that among the emotions a person can direct at a situation some fit it better than others.  The point can be extended: among the varieties of anger, or of sadness or exhilaration, that a person can direct at a situation some fit it better than others.  So accuracy cuts across our usual classifications for the emotions.  You can be miserable, elated, or curious, and be so in a way that does or does not accurately represent your situation.  No emotion is intrinsically accurate.  But some distinctions between emotions are necessary for a creature that is to have accurate emotions.  Sadness must be distinct from depression; remorse, guilt, shame, and embarrassment must be kept apart.  Falling into one of these when another fits the situation is a sure route to emotional mess.  And finding ones way around a rich range of emotions is as demanding as finding ones way around a complex set of beliefs.  It requires a special and admirable quality that it makes sense to call emotional intelligence.

            Emotional intelligence will not always result in emotional accuracy, any more than theoretical intelligence will always result in true belief.  And just as truth bears a complex relation to the coherence of belief, emotional accuracy bears an equally complex relation to the coherence of emotions, with one another and with a person’s complex of beliefs and desires.  Sometimes the more accurate emotions a person can have will not cohere well with one another or with the person’s other states.  This will typically be when the other states are defective, or when the situation is so complex that the person is not capable of coherent attitudes that represent it well.  (But then, the universe is like that, compared to our little brains.)  And, to pile on the warnings, there is no more guarantee that emotional accuracy will give us better lives than there is that we will be happier if we have true beliefs.  To the perspicacious tyrant who kills you if you don’t believe he is charming there corresponds the situation that is so unbearable that your sanity will not permit you to react to the way it really is.  But, we all trust, these are aberrant outlying cases.  In general, the route to truth leads through evidence and results in satisfied desire, and the route to emotional accuracy leads through the acquisition of a range of possible feelings and attitudes and the capacity to discriminate between them, and results in the harmony of thinking and feeling.  More specifically, it tends to link the evolution of our beliefs, our desires, and our feelings, and allows the present state of each of these to put pressure on the others.  It allows us to be whole people, by having patterns of thought that make two-way connections between what we believe and what we feel.  (Some of the connections in one way are clear: when you discover the insect is harmless your fear should change.  The connections the other way must consist in part of your emotions helping select relevance of evidence and direction of thought.  If you feel instinctively afraid of the insect you look for reasons, both in what you can see around you and in what you know, which might settle the question of its dangerousness.) 

            Imagine then a progression.  It starts with our hard-wired emotional responses, with their fixed affects and their simple paradigm scenarios.  Emotional intelligence then intervenes, and we acquire the capacity to modulate our emotions to what we learn and what we come to want.  (At the beginning we feel dismay at a situation; at a later stage we anticipate regret for the action we are choosing; at a yet later stage we anticipate regret if we take one choice and remorse if we take the other.)  Suppose that the capacity were perfectly acquired.  Then our emotions would match our situations to the extent that our information about then was accurate.  Could they then be counted as emotional truths?  The main factor to consider is the way they exclude one another.  At the original primitive stage fear, say, and delight are mutually exclusive.  And the exclusion is not just the effect of quirks and limitations: it is intrinsic to a simple fear that it leads one to intend avoidance and to a simple delight that it leads one to intend contact.  They are emotions that cannot both be held, though we can oscillate between them.  But each might be equally accurate.  As de Sousa, following Tappolet (2000), would put it, the values of danger and of attractiveness are both present.  So we shouldn’t count them as truths.  (It would be a strange kind of truth, such that having it committed one also to falsity.  To fear is to take as not attractive.)  But at later stages the exclusion lessens.  We acquire more subtle emotions, such as a delighted horror.  (You see the notorious association between sophistication and perversity.)  Then it is possible to acknowledge that the situation is both dangerous and attractive.  So as our emotions become more and more refined they come to be capable of representing more and more of the values present in our situations, in such a way that to acknowledge one is not to reject another.

            Might there be an ideal end to this progression, where in any situation an agent could have emotions which accurately represent it, and which do not exclude any others that accurately represent it?  I have no idea.  I do fear that these kinds of heroically accurate emotions would have become so much like beliefs that they could not easily serve the functions of emotions.  After all, as Greenspan and earlier work by de Sousa taught us, emotions are essential for defining patterns of salience that create pressures on the evolution of our beliefs and desires.  These patterns are essentially selective; they make things possible for us by limiting the possibilities.  But perhaps creatures with sufficient emotional intelligence would be able to assume these deliberately limiting perspectives while also remaining open to alternatives.  Perhaps.  We don’t have to take a position on this, in order to conclude that there is such a thing as emotional accuracy, that it is valuable, and that intelligent thinking and feeling aims at it. 

 

 

 

REFERENCES

 

de Sousa, Ronald,1987, The rationality of the emotions (Cambridge, MA: MIT Press).

Goldie, Peter, 2000, The emotions: a philosophical exploration (Oxford UK: Oxford University Press). 

Goldie, Peter ed.,2001, Understanding Emotions  (London: Ashgate).

Greenspan, Patricia, 1988, Emotions and reasons (London: Routledge).

Horwich, Paul 1998, Truth, second edition (Oxford UK: Oxford University Press).

Lewis, David, 2001, ‘Forget about the “correspondence theory of truth”’, Analysis 61: 275-280.

Tappolet, Christine, 2000, Emotions et valeurs (Paris: Presses universitaires de France).

Williams, C.J.F., 1976, What is truth? (Cambridge, UK: Cambridge University Press).