<<

. 56
( 82 .)



>>

SIMULATION, EMOTIONS, MORALITY 309
AND


Moreover, we do not just simulate the behavioral intentions of others.When
we deliberate about practical decisions, we simulate our own prospective activi-
ties.We imagine different lines of action and run them off-line on our own cog-
nitive/conative system in order to gain a sense of how we would react in different
circumstances as well as how we would feel emotionally about undertaking dif-
ferent lines of action.Thus, simulation is a crucial ingredient in practical delibera-
tion about our own actions.
In the case of simulating others, we input their relevant beliefs and desires into
the black box of our own off-line cognitive/conative system and then consider
the output as a predictor of their behavior. With respect to our own prospective
actions, we input our own beliefs about some possible future state into our off-
line or disengaged cognitive/conative system and contemplate our reactions,
including our emotional reactions, to alternative states of affairs. Obviously, from
an evolutionary point of view the capacity to run these off-line simulations is an
advantage. It is an economical way to figure out what others (including other
people and perhaps sometimes animals) will do. But we do not just simulate the
behavioral dispositions of others.We simulate our own prospective, future selves.
This enables us to test out alternative strategies in thought.
From the viewpoint of evolutionary theory, the explanation for why we have
the faculty of imagination/simulation is that it affords the capacity for strategy
testing.7 By entertaining thoughts about future states, we are able to get a handle
on how we will feel and act in alternative situations, and, as well, we are able to
work up informed hypotheses about how others are likely to respond to us
(which is useful in testing out our own prospective strategies). Simulation is a
means for constructing cost-free test runs of future actions that can provide us
with knowledge about ourselves and others.
Clearly, the theory that folk psychology is simulation rather than a complex
psychological theory avoids some of the problems of the Theory Theory.Accord-
ing to the simulation theory, there is no reason to hypothesize our dubious pos-
session of an immensely elaborate, subconscious psychological theory. Moreover,
the mobilization of such an elaborate theory with respect to particular cases
would seem to require a large amount of computing time, whereas running a sim-
ulation is a much faster process “ one whose speed is much more in keeping with
our actual, real time predictions of the behavior of others.
Moving from the realm of predicting actual behavior to the realm of aesthetics,
the application of simulation theory to the consumption of fiction is very straightfor-
ward.When we read, view, or listen to a fiction, we are running our cognitive/cona-
tive system off-line already; that is, we are imagining the story. Moreover, simulation
is a special case of imagining. It is, the simulation theorist argues, one of the primary
resources of the imagination that we employ when following texts.
With respect to fictional texts, Gregory Currie distinguishes between two
types of imagining. Primary imagining is a matter of what we may describe as
entertaining a proposition in the mind as unasserted. It is imagining, under the
guidance of the author, that such and such is the case in the world of the fiction.
310 ART, EMOTION, MORALITY
AND


But there is a another kind of imagining that Currie believes comes into play in
response to the fiction. It involves imagining “ that is, simulating “ the experience
of a character.
Currie writes:
Secondary imagining occurs when we imagine various things so as to imag-
ine what is fictional. Sometimes, secondary imaginings are not required for
primary imagining to take place: the story has it that a certain character
walked down a dark street, and we simply imagine that.Then we have pri-
mary imagining without secondary imagining. Primary imagining most
notably requires the support of secondary imagining in cases where what we
are primarily to imagine is the experience of character. If the dark street
hides something threatening, the character who walks may have thoughts,
anxieties, visual and auditory experiences and bodily sensations about which
it would be important for readers to imagine something. The author may
indicate to greater or lesser degree of specificity, what the character™s experi-
ence is. But it is notoriously difficult, and in some cases perhaps impossible,
for us to describe people™s mental states precisely.Authors who adopt stream
of consciousness and other subjective styles have failed to do it, and so have
film makers like Hitchcock who try to recreate the character™s visual experi-
ences on screen. Anyway, the attempt at full specificity and precision in this
regard would usually be regarded as a stylistic vice, leaving, as we significantly
say, “nothing to the imagination.”What the author explicitly says and what
can be inferred therefrom, will constrain our understanding of the character™s
mental state. It will set signposts and boundaries. But if these are all we have
to go on in a fiction, it will seem dull and lifeless. It is when we are able, in
imagination [through simulation], to feel as the character feels that fictions of
character take hold of us. It is this process of empathetic re-enactment which
I call secondary imagination.8
Simulation or secondary imagining, moreover, can be relevant to moral
deliberation.
We imagine ourselves in a certain situation which the fiction describes,
imagining ourselves to have the same relevant beliefs, desires and values as
the character whose situation it is. If our imagining goes well, it will tell us
something about how we would respond to the situation, and what it
would be like to experience it: a response and a phenomenology we can
then transfer to the character. That way we learn something about the
character. More importantly, from the point of view of moral knowledge,
we learn something about ourselves and about the things, we regard or
might regard as putative values.9
Fictions, by way of simulation, then, supply us with the kind of knowledge that
would be relevant to making a moral decision about a course of action “ knowledge
of what it would be like (e.g., what it would feel like emotionally) to be a liar, a
SIMULATION, EMOTIONS, MORALITY 311
AND


cheat, or a philanthropist. If we killed someone, could we live with ourselves? Sim-
ulation can provide some information toward answering such questions.
Moreover, this conception of the relation of fiction to morality provides a
means for evaluating narratives from an ethical point of view. For example, “fic-
tions that encourage secondary imaginings, while providing signposts for those
imaginings which systematically distort their outcomes, may do moral damage by
persuading us to value that which is not valuable.”10 And, presumably, fictions that
encourage us to value what is morally valuable are ceteris paribus, to be assessed
positively from the moral point of view.
There is, of course, a debate in the philosophy of mind about whether sim-
ulation is the correct conception of folk psychology.11 This is not a debate
about whether there is such a thing as simulation, but whether folk psychology
is basically a matter of simulation. But those arguments, interesting as they may
be, are not what we need to consider now. Rather, ours is the question of
whether the notion of simulation “ whether or not it best models folk psy-
chology “ is really relevant to aesthetics, specifically: is it relevant to our typical
intercourse (especially our emotional intercourse) with fictions and the moral
evaluation thereof.
For Currie, simulation is not the whole story of our engagement with fiction.
Nor does he claim that simulation is the only relation of narrative fiction to
morality. It is a relation, though one does have the impression that Currie thinks
that it is a rather central and comprehensive one. But is it really?
Simulation or secondary imagining, as Currie describes it, is not the same
thing as identification. For unlike identification, simulation does not presup-
pose that all of our cognitive and/or emotional states are identical to those of
the character whom we are simulating. As in everyday life, simulation only
requires rough similarity, not mental fusion. There is psychological evidence
that audiences do represent the emotional states of characters mentally.12 One
issue is whether that representation takes the form of simulation. At least one
psychologist has suggested that something like simulation might play a role in
understanding the emotions of fictional characters; but this has not yet been
substantiated empirically.13 However, supposing simulation sometimes comes
into play, the question is how often does this happen? How useful is simulation
theory as a comprehensive model of our commerce with fictions, especially
with reference to the emotions and morality?
Simulation theory suggests that we become engaged emotionally with fictions
by simulating the emotional states of characters. Our emotional responses to the
fiction are, so to speak, routed through the emotional states of characters. We
experience the fiction as if from inside the characters. But an alternative view
might suggest that our emotional responses to the fiction are more direct.We do
not typically emote with respect to fictions by simulating a character™s mental
state; rather, we might argue, we respond emotionally to the fiction from the out-
side. Our point of view is that of an observer of the situation and not, as simula-
tion theory suggests, that of the participant in the situation. When a character is
312 ART, EMOTION, MORALITY
AND


about to be ambushed, we feel fear for her; we do not imagine ourselves to be her
and then experience “her” fear.
In order to contemplate the differences between these two approaches, it is use-
ful to recall a distinction made by Richard Wollheim.With reference to the imagi-
nation,Wollheim distinguishes between central imagining and acentral imagining.14
Acentral imagining is a matter of my imagining that such and such; central imagin-
ing is a matter of my imagining x. Acentral imagination is exemplified by the case in
which I imagine that Kubla Khan built Xanadu; central imagining is exemplified by
the case where I imagine building Xanadu.Acentral imagining is from the outside,
so to speak; central imagining is from the inside. Given this rough distinction, Cur-
rie™s notion of simulation (or secondary imagining) is a case of central imagining,
whereas the alternative view stated in the preceding paragraph is involved with
acentral imagining.That is, according to the alternative view, we respond to fictional
situations as outside observers, assimilating our conception of the character™s mental
state into our overall response as a sort of onlooker with respect to the situation in
which the character finds himself. In contrast, for Currie, when we are involved in
simulation or secondary imagining, we are centrally imagining that we are the char-
acters.Which, if either, of these approaches is more comprehensive? Which models
our response to fictional narratives better?
I think that quite clearly as consumers of fictions we are typically in the position
of outside observers, or, as Richard Gerrig and Deborah Prentice call it, side-partici-
pants.15 Of course, the simulationist can respond that outside observers can employ
simulation. However, I wonder how often we do.After all, with most narratives, espe-
cially mass narratives, omniscient narrators tell us what is going on in the minds of
the characters. Simulation theory putatively informs us about how we go about pre-
dicting the behavior of others and understanding their affective states. But most nar-
ratives, it seems to me, give us ready access to the mental states “ the intentions,
desires, and emotions “ of characters. So what need do we have for simulation? We
have the information already in most cases. Furthermore, this often happens in visual
narratives as well; with respect to Casablanca, we do not have to simulate Rick™s feel-
ings about Ilsa; Rick tells us all we need to know in order to feel sorry for him.
Nor does this entail that the “direct access” to character™s inner states renders nar-
rative representations affectively lifeless. For we use that information, along with the
information about the situation the character is in, in order to generate our own emo-
tional reaction to the character and his or her circumstances.There is no need to sup-
pose that our affective state has to be channeled through a simulation of a character™s
putative state.We can generate our own emotional reaction directly (i.e., without an
intervening stage of simulation) by using the information that the narrator supplies us
about the character about whom we are concerned “ including explicitly given
information about her intentions, desires, emotions, plans, and so on.
There are two parts to this objection.The first is that with the typical narra-
tive, there is little role for simulation with respect to fictions “ especially written
fictions “ because the determination of what is going on in the mind of characters
is generally supplied by omniscient narrators.Thus, the pressure for philosophers
SIMULATION, EMOTIONS, MORALITY 313
AND


of art to use simulation to explain our grasp of a character™s state of mind does not
match up with the pressure for simulation theorists in the philosophy of mind to
explain our real-life predictions and understandings.
Second, as with the objections often advanced against the notion of identifica-
tion, simulation theory seems to overestimate the degree to which responding emo-
tionally to a fiction requires centrally imagining the states of characters. Most often, I
would contend, the emotionally appropriate object of our attention is the situation in
which a character finds herself and not the situation as the character experiences it.
The character feels grief, but we feel pity for her, in part, because she is feeling grief.
The object of her emotion is, say, her child.The object of our emotion is her situa-
tion “ a situation in which she is feeling sorrow.We do not simulate her situation;
rather, we respond emotionally with our own (different) feeling of pity to a situation
in which someone, namely the relevant character, is feeling sorrow.
Putting these two objections together, then, we can argue that typically we do
not need to postulate the operation of simulation because our emotional response
is finally that of an observer (not a direct participant, as simulation might suggest),
and the relevant information needed to form the appropriate emotional response
from an observer™s point of view is generally supplied by omniscient narrators.
Thus, there is no reason to postulate the operation of simulation in the typical case
of responding to fiction.
Of course, this argument, if it is persuasive, might be thought to apply primar-
ily to written and perhaps spoken narratives. It might be said that it is less com-
pelling when it comes to visual narratives like movies and TV programs. For with
visual narratives, it is far less customary to have the sort of omniscient narration
where we are given direct access to the minds of the characters. Running
voiceover commentary on the characters™ internal states “ in either the first person
or the third person “ is, for example, rare. So, it might be argued that in general
when it comes to visual narration, simulation usually has a role to play of a sort
that I have denied it plays with standard cases of written narration. In visual nar-
ration, we are given the character™s overt behavior and have to go on from there.
Might it not be the case that we go on by way of simulation?
My inclination is to resist this suggestion. First of all, as the example of
Casablanca indicates, characters often tell us about their mental states “ their inten-
tions and their feelings “ outright. But in addition, once again there is what we
may call an asymmetry problem in typical cases of narration.Typically our emo-
tional responses to characters are different from their emotional responses.We are
paralyzed by fear when the heroine is trapped by pursuers on the edge of a para-
pet, but she, undaunted and fearless, plunges into the moat several hundred feet
below.16 We feel sorrow for characters wracked with guilt.That is, the emotion we
feel is different from the emotion felt by the character. There is no symmetry
between our feelings and the character™s feeling “ which is what simulation theory
would predict. Rather, there is asymmetry, at least in a large number of standard
cases. And in those cases, simulation just doesn™t seem to be the right model for
these audience responses.
314 ART, EMOTION, MORALITY
AND


But how, it might be asked, do we know the characters are wracked with guilt,
since without knowing that, we may not respond with pity? Don™t we need sim-
ulation to explain this? I think not, and not only because characters often verbal-
ize their internal states. Rather, we can recognize the states of others without
simulation.17 This is not a reversion to the Theory Theory. Rather, we need only
suppose that people have the power to recognize certain patterns. This does not
require having a full-blown theory, but only a repertory of sometimes related,
sometimes unrelated schemas or prototypes for assessing situations. For example,
in order to interpret the emotions underlying a convicted criminal™s effort to
shield his face from a TV camera, I need not simulate his mental state in order to
recognize that it is connected to his sense of shame. Likewise, tracking the emo-
tional states of characters in films and TV programs rarely requires simulation. It is
easy to recognize their states without simulating them.18
For example, consider point-of-view editing. It might be suggested that that is
a form of visual narration that must involve simulation. The character looks off-
screen, and in the succeeding shot we see molten lava streaming toward the cam-
era. Don™t we feel fear because we are simulating the character™s response? I don™t
think so.We know that molten lava is dangerous without imagining ourselves to
be in the character™s position. If we are concerned about the character, the knowl-
edge that molten lava is heading her way is enough to engender fear for the char-
acter in us. The added step of imagining that we are in the character™s shoes is
unnecessary.19
Moreover, we can confirm that this is enough to explain our response by not-
ing that our fear for the character may be no different whether we suppose the
character knows she is about to be engulfed in lava or not. Presumably, the simu-
lation theory would predict different responses to these alternative situations, since
we would be simulating different mental states. But I suspect that we can vary the
mental states of the characters without provoking a difference in our emotional

<<

. 56
( 82 .)



>>