We have differentiated between the thought experiments carried out by professional philosophers with their colleagues in mind and the “democratized” thought experiment.
More concretely, a “democratized” thought experiment unfolds as follows:
1. We present to subjects selected according to various criteria deemed to be pertinent (young or adult, boys or girls, educated or not, religious or not, and so on) a range of little fictions that are supposed to arouse their moral perplexity, such as The Killer Trolley. They are either presented in written form as “vignettes” (to use the technical term), or else recounted by the experimenter. They end with questions such as “what would you do?,” “what must be done?,” “did he do well?,” “is it morally permissible?,” and so on.
2. We note the spontaneous answers of the people exposed to the narrative.
3. We ask them to justify their spontaneous judgments.
4. We seek to account for the statistical distribution of the answers.
5. We attempt to draw more general conclusions as to the validity of moral theories: consequentialism, deontologism, virtue ethics.
The “democratized” thought experiment, in moral philosophy, consists in this whole sequence: construction of the moral fiction, presentation to the largest possible sample selected according to the most varied criteria, recording of the spontaneous judgments and a discussion of the attempts to justify such judgments, the comparing of the explanations with the causes and the reasons, and theoretical conclusions.
The thought experiment for philosophers skips the second and third stages: presentation to the largest possible sample selected according to the most varied criteria and recording of the spontaneous judgments and a discussion of the attempts to justify such judgments.
All these experiments concern our moral beliefs, that is to say, what we find good or bad, desirable or undesirable, just or unjust, whether these beliefs are spontaneous or reflective.
They enable us to evaluate the validity of consequentialist or deontological intuitions, as well as to rethink one of the most traditional questions in moral philosophy: does there exist an innate, universal “moral sense,” a “moral instinct,” and exactly what form does it take in our minds?
We have also distinguished between these thought experiments and behavior experiments: helping or destructive behaviors as the case may be.
Experimental moral philosophy has taken a particular interest in laboratory experiments such as that of Stanley Milgram, with a view to putting the idea of a moral or immoral “personality” to the test.
When we are faced with this huge program, one objection may immediately spring to mind. Is it not a particularly naive project, insofar as it seems to place its trust in research that raises methodological and epistemological problems that are far from negligible?
METHODOLOGICAL PROBLEMS
In the majority of works to which moral philosophy refers, the subjects of the investigation are exposed to imaginary situations, which they have probably never faced in their lives, at any rate in this simplified form.
These works suffer from the shortcomings characteristic of the genre: the difficulty of evaluating the exact scope of the results outside of the experimental conditions, the tendency to force the subjects’ answers into preestablished categories that are perhaps not their own, and so on.
Furthermore, the conclusions of this research are formulated in statistical terms, which raises all manner of problems regarding the number of subjects involved in the experiment (which is sometimes not so numerous), their “representativity,” and the thresholds beyond which a result is held to be
meaningful.1If a piece of research claims to establish that “evil is banal” because 20 percent of thirty students in moral philosophy were prepared to administer painful electric shocks to their peers in the context of a paid experiment, one would not be wrong to mistrust it.
EPISTEMOLOGICAL PROBLEMS
All this research (laboratory experiments and investigations in the field) belongs to the human sciences. Now the possibility (as well as the intrinsic interest) of aligning these disciplines with the natural sciences, by using their methods, with the same explicatory and predictive ambition, is still highly controversial.
The more skeptical reckon that the project of applying, to human behavior in general and to psychological states in particular, methods that have only been tested in the explication and prediction of physical events has no chance of success.
2If the human sciences are not, and never will be, rigorous sciences, explicatory and predictive, even when they imitate the latter’s methods, do they amount to anything more than armchair philosophy? What is the point of wasting time in examining them more closely?
One should add that, for a good many philosophers, the vocation of the human sciences is such that the interest of their results, supposing there were any, would be far from evident for moral reflection.
The human sciences seek to inform us about what is. They describe facts. One of the vocations of moral philosophy is to tell us what is good, what we should do. It proposes norms.
Now, say these philosophers, we cannot derive any norm from a simple fact. Thus, from the fact that the majority of people give nothing to organizations combating famine, it does not follow that this is good or what we should do.
These methodological and epistemological objections are well known to philosophers, who have taken the wager of interesting themselves in empirical research. They are generally very aware of methodological problems. The laboratory experiments and field investigations they use are few in number. Such works serve as reference points, having withstood constant methodological criticism since they were first published.
3The two epistemological objections appear to be more difficult. But in reality they are not conclusive, given the limited aims of experimental moral philosophy.
In order to determine what is just or unjust, good or evil, moral reflection cannot do without references to the moral intuitions of each and every one of us, any more than it can do without references to typically human “capacities” or “needs”
4In order to analyze moral judgments or behaviors, moral reflection does indeed have to refer to peoples’ “motivations,” “intentions,” “emotions” “character” and “personality”
In other words, moral reflection is never wholly independent of certain facts, in the sense that it invariably uses concepts whose characterization is linked to certain facts (the moral intuitions of each and every one of us, typically human needs, peoples’ motivation and character, and so on).
“Experimental” philosophers only intervene when these concepts are deployed in their colleagues’ arguments.
They question the privilege that certain philosophers accord themselves of thinking that they know more than anyone else about such concepts, without having taken the trouble to go and look at them.
But this does not mean that “experimental” philosophers wholly reject the idea that the passage from facts to norms poses problems, or that they naively disregard the limits of the human sciences.
