- seCtIon FoUR: eVoLUtIon
Povinelli (1998) agrees that chimpanzees have a concept of self, but not that
they are aware of their own psychological states (for the counter-argument, see
Gallup, 1998). He suggests that ‘self-recognition in chimpanzees and human tod-
dlers is based on a recognition of the self ’s behaviour not the self ’s psychological
states’ (p. 74). Others have argued that MSR is a mere by-product of a more gen-
eral ability to collate and compare multiple mental models of the same thing – a
skill demonstrated in search tasks and pretence as well as the ability to build and
update expectations about their own physical appearance (Suddendorf and But-
ler, 2013). Even more sceptical is British psychologist Cecilia Heyes (1998), who
agrees that chimps are capable of ‘mirror-guided body inspection’ (p. 102), but
argues that they have no self-concept and no understanding of mental states. For
her, ‘mirror self-recognition’ is the wrong way to describe the test; all it indicates,
she says, is one form of mirror-guided exploration. But others have suggested
that she takes such a hardline stance because she assumes that self-recognition
is an all-or-nothing capacity which depends on being able to form second-order
representations of oneself, rather than it being possible to recognise oneself in a
more naïve way (Brandl, 2016).
KNOWING THAT OTHER MINDS EXIST
We humans have beliefs, desires, fears, and intentions, and we attribute them to
others. That is, we have a ‘theory of mind’, or can ‘mindread’ or ‘mentalise’.
Early theories of social cognition proposed that on the basis of what other peo-
ple say, how they look, and what they do, I can infer their (unobservable) mental
states, and on that basis I work out how to interact appropriately with them.
This model, in which I construct a theory about the causes of other people’s
behaviour, is now often referred to by the shorthand of ‘theory theory’, and has
more recently been challenged by ‘simulation theory’.
In simulation theory, I understand others by running a simulation of their
actions, as if performing them myself. The simulations can be conceived of as
either conscious (or ‘person-level’, or ‘explicit’) or unconscious (or ‘sub-personal’,
or neural, or ‘implicit’). Either way, a simulation generates pretend versions of
the other person’s states in me, which allows me to grasp their thoughts, beliefs,
and desires.
More recently still, ‘interaction theory’ has proposed that there is no need to posit
any indirect, mental route to calculating someone else’s ‘inner state’, whether by
inference or simulation. On the contrary, I understand other people as I do myself,
as an embodied agent in constant interaction with others, within a constraining
and affording environment. To see someone smile is not a piece of evidence to
feed into an inferential calculus or a mental simulator, but a direct experience
of their pleasure or happiness. ‘Accessing your thoughts, beliefs and desires thus
becomes, for Interaction Theory, less a matter of reading your mind than attend-
ing to the world we already share’ (Chesters, 2014, p. 71).
Possibly all three theories can help us understand how social cognition operates in
different circumstances and responds to different challenges. Interaction theory
certainly offers more scope for non-human interactions to qualify as fully fledged
social cognition. All three, however, are compatible with Dennett’s (1987) notion
‘Can animals
empathize? Maybe not.’
(Povinelli, 1998)
‘Children and chimps
and crows and
octopuses are ultimately
so interesting not
because they are
mini-mes, but because
they are aliens – not
because they are smart
like us, but because
they are smart in
ways we haven’t even
considered.’
(Gopnik, 2016)