60
DAVID T. LYKKEN AND AUKE TELLEGEN
but they do not explain why he chooses one dish rather than
another when the time comes.
Choices. Buying a house, a car, or a suit involves something
analogous to the just deserts idea in that we know what we can
afford to pay and we confine our search to items in that price
range. What we do then is to rank order the possibilities accord-
ing to subjective criteria and then select the top ranked item.
The similarity model might be taken to imply that, in selecting
a mate, we rank order available candidates within our search
area according to their similarity to ourselves and thus accom-
plish a specific selection. Yet, there seems to be no evidence at
all that we do this and, moreover, if we did, the spousal correla-
tions would be higher than in fact they are. That is, if we restrict
the field of eligibles to people whose absolute difference from
ourselves in, say, physical attractiveness is 0.8 standard devia-
tions, then the resulting spousal correlation in physical attrac-
tiveness will be about SO. But, if we proceed to choose among
available eligibles that person closest to ourselves in physical
attraction, then the spousal correlation would be substantially
higher than .50. Therefore, this crucial second step of selecting
by a paired comparison process the best of our short list of
eligibles, the step that characterizes most choice behavior and
makes singular selection possible, does not seem to occur in the
selection of mates, not at least if we are selecting on the basis of
the same standard set of nomothetic dimensions. The similar-
ity data indicate that we exclude candidates too different from
ourselves, but we know of no evidence that, given several candi-
dates, we tend reliably to select that individual among them
who is closest to us in the several-dimensional similarity space.
Especially because such selection would yield correlations
higher than those observed, then absent such evidence, it ap-
pears that the similarity model (and the just deserts model)
must be rejected as a theory of specific mate selection.
Poverty or last chance model. It could be argued that the
discussion so far, which seems to imply that mate selection is
like choosing clothes off a rack or a meal in a cafeteria, is unreal-
istic; most of us do not have that many choices. As Kerckhoff
(1974) pointed out, many of us only know people in our neigh-
borhood, school, church, or workplace, and most of those peo-
ple resemble us enough, because ofthe ethnic and social stratifi-
cation that these institutions naturally impose, to fit within our
similarity hypercubes. By thus limiting our field of availables,
propinquity alone might be responsible for the spousal similar-
ity that we observe. Perhaps most of us do not do any selecting
at all but, instead, accept the first serious prospect that comes
along during the period(s) in our lives when we are interested in
getting married. This poverty or “bird-in-the-hand” model sug-
gests that mating is like the pairing off of couples at a dance or
at a singles bar; he asks because she’s there and looks to be
available and she accepts because he is the only one who’s
asked her.
Unlike the similarity and the just deserts models, the poverty
model does not leave the seeker in the kosher line, unable to
decide, because this model tells us that there will be only one
entree available. It can hardly be doubted that something like
the poverty model reasonably describes at least some marriages
today, nor is there compelling reason to doubt that many such
marriages work out very well. Most marriages, in many parts of
the world, are arranged alliances in which the participants have
little or no power of choice and, therefore, are similar to the
single-option situation of the poverty model, and we must pre-
sume that most such arranged marriages successfully accom-
plish the primary purpose of marriage: the bearing and rearing
of children.
Indeed, it seems likely that the poverty model reasonably
describes the situation that prevailed during the Pleistocene
era, when our ancestors lived in small bands and between-band
contact was limited and precarious. It may well be that circum-
stances permitting a real range of mate choice are so recent in
human history that an adaptive mechanism for mate selection
has not been well shaped by evolutionary pressures. We shall
argue, however, that our species has evolved a mechanism for
pair bonding and that it is this mechanism which, for many in
modern Western societies, creates an illusion of mate choice.
But let us meanwhile accept the common belief that most of
us in modern Western societies “marry for love,” that we do
have options and we do make choices, and see where it leads us.
As we have shown, the similarity model (which we shall hereaf-
ter assume to include the just deserts model as a component
part), is ^curate descriptively; spousal pairs do tend to resem-
ble each other in age, traditionalism, physical attractiveness,
education, stature and, weakly, in most other respects. This is
presumably because most of us tend to acquire mates from
among candidates, most of whom resemble us in these ways,
either because we restrict our search to this group or because
this is the only group available to us because of the effects of
social stratification. But the similarity model has not been
shown to be able to account for our actual choice among that (in
principle, large) group of candidates.
Study 2: The Idiographic Model of Mate Selection
Searching for evidence of spousal similarity-positive assor-
tative mating-is methodologically easy, a matter of assessing
married couples on a variety of traits and then computing Spou-
sal correlations. But this approach is reminiscent of the ine-
briate who has lost his car keys in a darkened doorway but
elects to look for them under the street light. It would be hard to
find, in romantic poetry or in real life, a description of a be-
loved that is couched in terms of similarity to the self-descrip-
tion of the lover. The obverse hypothesis, negative assortative
mating or the attraction of opposites, is an ancient idea re-
flected in the belief that one seeks in a mate the other half of the
perfect being of whom the first half is one’s self. But this comple-
mentary model of mate selection is invalidated by the same
weak but consistently positive spousal correlations already
mentioned.
A more likely hypothesis is that each individual has a unique
set of criteria, so that we are not all competing for the same few
“perfect” mates, and that any random John or Marsha will use
idiosyncratic criteria that may include some similarities, some
differences, and even some opposites. Winch’s (1958) theory of
complementary needs and Murstein’s (1976) stimulus-value-
role theory of marital choice are variants of this idiographic
model of mate selection.
Similarity of the Mates of Monozygotic Twins
The idiographic model of mate selection asserts that each
individual possesses certain reasonably discriminating criteria