also been shown to be associated with both the “hour-
glass” shape of a female’s body (a 0.7 waist to hip ratio)
and higher than average fecundity.In the absence of
contraception, female fecundity reaches its maxi-
mum in the mid-20s, declines by about 20% in the
mid-30s, and then falls precipitously by a further
60% during the 40s. The thinning of a female’s lips
parallels these steep declines in fecundity and it is not
uncommon for females to use lipstick or collagen in-
jections to maintain or enhance their facial attractive-
ness. Taken together, these observations suggest that
the physical beauty of a human female’s face depends
on the display of highly visible hormonal markers
that are indicative of higher than average fecundity.
Any facial characteristic that enhances biological
survival will increase in frequency over generations as
a consequence of natural selection. For example, in a
cold dry climate a long nose with thin nostrils that can
warm and add moisture to the air would be an adap-
tive trait, whereas a short nose with wide nostrils
would be adaptive in a tropical environment. So in
any breeding population these average facial features
should evolve over generations because they are the
most adaptive traits for survival in the local environ-
ment. Sexual selection is a form of natural selection,
but it has a different modus operandi. In sexually re-
producing species, sexually selected traits are attrib-
utes of one sex that enhance access to the opposite sex
or offer a competitive advantage over members of the
same sex. These traits are mediated by sex hormones
and can be in the form of bird songs or conspicuous
tails or tusks that are displayed by members of one sex
or the other. Viewed from this perspective, in any
local breeding population the average female face is
certainly attractive but the most attractive faces dis-
play the additional hormone mediated sexually se-
lected traits that serve as markers of higher than
average fecundity (see Figure 1). Such displays would
be ineffective, however, if they didn’t evoke an emo-
tional response in the brains of male admirers. This
hypothesis has been evaluated by recording event re-
lated potentials(ERPs) from human males exposed to
brief presentations of facial images of the opposite sex.
ERPs are obtained by measuring the changes in
voltage recorded from different scalp electrodes during
the one-second interval following the presentation of a
stimulus. When the waveforms collected from each
electrode are averaged over a number of presentations
of the same stimulus the averaged ERP reveals the
spatial and temporal pattern of the neural processes
evoked by that stimulus. The averaging procedure en-
hances the consistent effects evoked by the stimulus
relative to the variable background activity, thus im-proving the signal to noise ratio. The early components
of ERPs vary with the physical attributes of a stimulus,
such its intensity, color, or modality, but the later com-
ponents are correlated with more general attributes
such as the probability of the stimulus or its value to
the observer.
One of the most studied late components is the
third positive wave (P3) that is characterized by its
maximum amplitude in parietal electrodes and a la-
tency of 300 to 600 milliseconds, depending on the
complexity of the stimulus. It has been well established
that low probability stimuli evoke larger P3 waves than
high probability events. That is, if an improbable low
pitch tone is embedded within a long sequence of
identical high pitch tones (an odd ball paradigm) the
P3 evoked by the low probability tone is much larger
than that evoked by high probability tone. However,
when subjects are presented with a random sequence
of two equally probable stimuli (tone or light) but
asked to predict which one will occur on any trial, the
P3 is always larger to the stimulus that disconfirms,
rather than confirms, their prediction. That is, the am-
plitude of the P3 is really a function of the observer’s
subjective probability, not the computed objective
probability. Simply stated, surprising events evoke a
larger P3 than expected events.
A second factor that influences the P3 amplitude
is the value of the stimulus to the observer. Cognitive
psychologists define the value of a stimulus as its “task
relevance” whereas biological psychologists define
value as the inherent or learned “emotional value” of a
stimulus to the observer. In my “Darwinian Goggles”
article I suggested that pleasant and unpleasant feelings
are an evolved inherent value system that defines exter-
nal objects or events as threats or benefits to reproduc-
tive success. Objects, such as money, can acquire value
by virtue of their learned association with these funda-
mental rewards and deterrents. This interpretation of
value has been supported by a number of experiments
that systematically examinedERPs evoked by biologi-
cally-relevant stimuli that consistently evoked pleasant
or unpleasant feelings in human observers.
As a first step, this experiment involved rating a
large number of pictures on a five-point scale that
ranged from very unpleasant to very pleasant. The
ratings were then used to construct three non-over-
lapping categories (most pleasant, neutral, most un-
pleasant), with 4 slides in each category. The content
of these categories was babies, ordinary people, and
dermatological slides that displayed faces with severe
skin diseases. Two additional four-slide categories
were constructed based on pictures that received dif-
ferent pleasantness ratings from males and females.2 0 S K E P T I C M A G A Z I N E volume 25 number 1 2020