MBB Junior Symposium – 15 September 2005


Simon Baron-Cohen¸ The Essential Difference: Male and Female Brains and the Truth about Autism. Basic Books: 2003.


Some definitions and theses from Baron-Cohen (chapter 1)

Male brain  (=Type S): A brain in which ‘systemizing’ stronger than ‘empathizing’.

Female brain (=Type E): A brain in which ‘empathizing’ stronger than ‘systemizing’.


NB: Baron-Cohen use the terms ‘male brain’ and ‘female brain’ to describe something other than brains of most males and females. It is not even part of his official theses that this is so. It is thus a bit misleading that there are chapters called ‘The female brain as empathizer: the evidence’, since here ‘the female brain’ is being used to refer to brains of actual females.


            Three central theses of the book

1. “Males spontaneously systemize to a greater degree than do females”, on statistical average. I.e., more males than females have brains of type S.

à Notice that it doesn’t follow from this thesis that most males overall have brains of type S. For all the thesis says, it could turn out that 90% of women have type S brains but 95% of men do.


2. “Females spontaneously empathize to a greater degree than do males”, on statistical average. I.e., more females than males have brains of type E.

à Notice that it doesn’t follow from this thesis that most females overall have brains of type E. For all the thesis says, it could be that 90% of men have type E brains but 95% of women do.


3. Autists have extreme male brain:

            Systematizing is much more developed than Empathizing than it is in non-



Some characteristics of autists (chapter 10)

·            Fine discrimination/Attention to details normally overlooked (e.g., preference for blanket A seemingly identical to non-preferred blanket B)

·            Memory (e.g., knows which car on street goes with which house and when parking permit expires)

·            Sensitive to patterned information (e.g., can recite railway timetable information).

·            Mathematical ability (e.g., fast at mathematical and calendrical calculation, is a number prime and if not, what are its factors)



“Empathizing is the drive to identify another person’s emotions and thoughts, and to respond to them with an appropriate emotion. Empathizing does not entail just the cold calculation of what someone else thinks and feels…Psychopaths can do that much. Empathizing…is done in order to understand another person, to predict their behavior, and connect or resonate with the emotionally.” (p. 2)

NB: by definition, you empathize with someone else only if their feelings affect your own.  (p. 5)

“Empathizing is about spontaneously and naturally tuning into the other person’s thought and feelings, whatever these might be” (p. 21).


Some examples of empathizing offered by Baron-Cohen  (chapter 3)

·            ‘constantly search people’s tone of voice and scan people’s faces, especially their eyes, to se what they might be thinking’.

·            ‘empathy makes real communication possible…it leads you to ask the listener how they feel and check if they want to enter the dialogue, or what they think about the topic’

·            ‘Imagine that you witness a crash..empathy propels you to sit with the victims of the crash, checking how they are, reassuring them that someone is there for them.’



“Systemizing is the drive to analyze, explore, and construct a system. The systemizer intuitively figures out how things work, or extracts the underlying rules that govern the behavior of a system. This is done in order to understand and predict the system, or to invent a new one” (p. 3)

            What is a system?  ‘They all operate on inputs and deliver outputs using ‘if-then’            correlational rules’ (p. 3)


‘The big pay-off of systemizing is control…Systemizing is an inductive process. You watch what happens each time you ___ , an after a series of reliably predictable results, you form your rule.’ (p. 67)


Some examples of systemizing offered by Baron-Cohen (chapter 5)


(**) We have probably all had an experience of doing something with one set of expectations, or with a rather unspecific set of them, and ending up surprised at the outcome. This would be a paradigmatic un-systematic outlook.



Baron-Cohen is aiming at a theory of behavioral that males and females supposedly manifest. The categories of systemizing and empathizing are supposed to help us understand those differences. The form of the explanation for the behavioral difference is then supposed go like this:


            (1) Overall, males and females differ in typical behavior.

(2) Males [/females] on the whole have stronger abilities to systemize [/empathize] than do females [/males].

(3) We would expect people with S-type brains to behave the way males typically do, and we would expect people with E-type brains to behave the way females typically do.

So (4) that’s a reason to think those who exhibit typical male behavior have S-

type brains and those who exhibit typical female behavior have E-type brains.


A problem for this explanation

It is very hard to avoid the drive to systemize, and very implausible to suppose that anyone exhibits even relatively lower levels of such behavior than anyone else. That is because the definition of systemizing encompasses a great deal more than Baron-Cohen acknowledges (and perhaps more than he intended the category to capture).


Other examples of systemizing that fit the official definition


Ordinary skilled activity

Artistic or aesthetic examples

Social interaction examples


…in short, any familiar goal-directed activity will be an example of systemizing!


Question: Does the expectation of the regular effect have to be available to consciousness? You might think that the examples above are done without awareness of the ‘if-then correlations’.


Answer: The case of autists suggests that the expectations are not available to consciousness even in the paradigmatic case of systemizing. E.g. autists can’t tell you how they calculate days of week for dates far into the future. Ditto for Baron-Cohen’s examples: the boy can’t tell you how he keeps track of information about the camermen in the movies.


à So it looks like skills of  empathizing are also skills of systemizing, by Baron-Cohen’s definition. Empathizers might ‘naturally’ modulate tone of voice without being aware that they are doing so, or of the correlations between tone and effect. But this won’t disqualify the activity as an instance of systemizing, if autists calculations, which are equally intuitive and spontaneous.


à To get a contrast with systemizing given B-C’s definition, you would have any expectations about the outcome of any action. Nearly all of their actions would be unthinking double-starred kind describe above.


Further question to think about


Ø        What exactly is meant by “spontaneously systemize/empathize” in Baron-Cohen’s main  theses:


Is the point that S-types [/E-types] involuntarily systemize [/empathize]? That systemizing [/empathizing] behavior comes very easily to them to do do? That they do so as a default, where this might be overridden by other (possibly involuntary) tendencies? … or by voluntary choices? Or that these are overriding tendencies?

o       In order for the claim about what’s done ‘spontaneously’ to explain the typical behavior, it seems as if it can’t be simply a default tendency, or something that comes easily, unless Baron-Cohen is prepared to defend the additional assumption that overall people do what comes easily.

o       It looks as if the strongest explanation for actual behavior would be that the drive to systemize [/empathize] is that this is behavior that one ends up doing no matter what else one tries or wants to do. That would make these tendencies relatively immalleable.

o       Offhand it doesn’t seem as if such tendencies are immalleable, in the case of normals.


Ø        Is the fact that someone is strong as empathizing, as measured by the tests Baron-Cohen offers that are supposed to operationalize it, supposed to itself be a piece of evidence that they are weak at systemizing? To the extent that the tests for empathizing are supposed to reflect ordinary notion of empathizing, it would seem as if this shouldn’t count as such evidence. (Baron-Cohen seems to think that the tests do reflect the ordinary notion, or at least aim to, since he says at the start of chapter 3 that if the reader knows what empathizing is they can skip the chapter). But in some place Baron-Cohen suggests that it would count as such evidence. Eg on p. 5 he says that to systemize one needs to be detached from the subject-matter being systemized, whereas to empathize one needs to be attached to the people involved. So he seems to think of each of these supposed abilities as limiting each other. And that seems false, at least for some of the supposed paradigms of systemizing and empathizing. Nothing about the intrinsic nature of mathematical ability, for instance, rules out having one’s feelings influenced by those of another.



n         Given how little it takes to count as systemizing, it would seem that to empathize one must also systemize. If so, then the claim that some people are very strong at empathizing and weak at systemizing is incoherent. The converse claim however, does not hold: one could (by Baron-Cohen’s definitions) systemize without empathizing, and therefore in principle one could be strong as systemizing without being strong as empathizing.


n         Perhaps there is a more finely-grained category of ability than systemizing that could explain behavioral differences in normals that Baron-Cohen sets out to explain. If not, this explanation does not seem very promising.


n         Suppose it turned out that males scored higher than females on the tests that were supposed to operationalize systemizing, and females score higher than males on the tests that are supposed to operationalize empathizing. What should we conclude from that? Well, given that if you empathize, you thereby also systemize, then if someone scores high on empathizing but low on systemizing, then the tests cannot each be operationalizing these abilities.  Maybe the test for systemizing is zeroing in on something else – a more narrowly defined set of abilities, or a set of interests.