"Mankind," sex marking and kin terms

Fri, 6 Jan 1995 08:11:00 PST

The sometimes virulent discussion of the word "mankind" led me to
look up its meaning in my Oxford Universal Dictionary. I was
surprised to find that what we often take to be its meaning;
i.e., a word which refers exclusively to the male sex, is not its
original meaning. The Oxford Dictionary dates the word back to
Middle English and provides two current meanings for the word.
The prior meaning is stated to be: "The human species" and as
illustration, "Mankind never suffer any work to be lost which
tends to make them more wise or happy" (Goldensmith). The second
meaning, which dates to 1526 is: "The male sex; persons of the
male sex" with illustration "The ... silliness of mankind and
womankind at large."

My Webster's New World Dictionary (1957 edition) gives: "1. all
human beings; the human race. 2. all human males; the male

I take it that the word "mankind" can be properly used to refer
to "the human race" as a gender neutral term; yet as with any
word which is polysemic there is the potential that one person's
intended meaning may be read according to another meaning of the
word. This would appear to be particularly problematic with a
word whose two significata have the relationship of set and
subset to each other, as is the case with the word "mankind;"
e.g., in the posting which began this long diatribe the word
"mankind" was used in a phrase where either meaning of the word
"mankind" could be substituted and have a meaningful phrase,
hence the potential for choice on the part of the reader as to
which meaning should be inferred as the intent of the author.

Several postings have directly or indirectly implied that perhaps
it is the existence of gender marked pronouns that leads to
exclusivity of reference and implied, if not actually stated, that
this leads to hierarchy in the assumed relationship between female and male
referents. This led to the suggestion that languages (if any) without
gender marked pronouns might thereby have a more equal viewpoint of the two
sexes, female and male.

I suspect that this is both naive and wishful thinking. The
underlying assumption is that if a dichotomy is made; e.g.
female/male, it necessarily involves hierarchy:


and with hierarchy, relationships of power/non-power,
control/controlled, etc. I suggest, following arguments made by
the eminent structural anthropologist Fadwa El Guindi, that splitting a
category into a binary opposition does not, in and of itself,
require hierarchy; that is, the transformation {person} -->
{female person, male person} does not simultaneously require the
structural relationship of hierarchy. Alternatively one can have
the structure:

/ \
/ \
female male

with the categories {female} and {male} coequal.

It seems to be the case that in our culture we have a
hierarchical model of power relationships:



which we map onto binary oppositions to give them a hierarchical

dominator ----> male

dominated -----> female

thereby arriving at:




with that hierarchical relationship taking on the properties of
the model which gives it structure; i.e., in our culture males
are viewed as dominating females, having power over females, etc.

This is not a universal structure; e.g. see seminal ethnographic work of
El Guindi on female/male relationships in middle eastern cultures, though it
is ubiquitous in our own culture; e.g. boss/worker, white/black,
teacher/student, etc.

In a previous post (many months ago) I suggested that much of
what we see in our culture can be viewed as attempts to remove the
gender distinction to achieve equality (i.e., to
erase the hierarchical relationship by denying the distinction upon which it
is based). Thus if we eliminate the original opposition:

{person} --X-> {female, male}

then the hierarchy does not exist (or so the argument goes).
This would account, for example, for the emphasis in our
culture on DENYING sex difference through symbolic means such as
clothing, or through alleged properties of the respective sexes,
etc. as a way to achieve equality, whereas the struggle for
equality between the sexes in middle eastern cultures takes on a
different dimension and there is no interest whatsoever in the
kind of denial of sex differences in clothing, difference in the
sexes, etc. that is very much the focus in our culture.

As for the deeper "location" of sex markings in language, let me
briefly indicate some differences that my work on kinship
terminology structure has elucidated. In our kinship terminology
we have terms of reference that are sex marked and terms that are
not sex marked. In some cases we have unmarked terms that serve
as cover terms for sex marked terms; e.g., Parent = {Mother,
Father}. Other terminologies have very different patterns for
the sex marking of terms. From what do these differences in sex
marking arise? The analysis I have done demonstrates that there
are fundamental differences in terminologies as to the structural
level (in terms of deep to surface structure) at which sex
markings come into play. One situation is where sex markings
arise not at the level of the generating terms for the
terminology, but as a rule applied to an already generated
structure. For example, in our own terminology, the deep
structure is based on non-sex marked terms: Parent, Child, Self.
Sex markings structurally arise only after a consanguineal-
affinal structure is generated and arises via the following rule:

"A kin term K will be sex marked when Spouse of K is a kin term
(i.e., is a kin term in the structure that has been constructed;
this is exemplified by Spouse of Parent = Parent as kin terms--
not kin types!--and so Parent --> {Mother, Father}) or if for the
reciprocal term, K', Spouse of K' is a kin term (for example,
Spouse of Grandchild is not a kin term, but Grandparent is the
reciprocal of Grandchild and Spouse of Grandparent = Grandparent,
so Grandchild --> {Granddaughter, Grandson})."

In contrast, other terminologies; e.g., the Shipibo, the
Trobriand and others, use as generators sex-marked terms. Thus
the terms used as generators for the Shipibo terminology are:
"Father", "Son", "Male Self" or, isomorphically, the generators
are "Mother", "Daughter", "Female Self". For the Trobriand the
generating terms are "Father", "Older Brother" or,
isomorphically, "Mother", "Older Sister".

Open question: Do these structural differences in how sex
markings occur within the kinship terminology relate in any way
to emic models of sex relationships?; that is, our terminology is
based upon the mapping Parent --> {Mother, Father} which
parallels the mapping Person --> {Female Person, Male Person} so
that our cognitive framework seems to allow for sex markings as a
derived property, hence more amenable (?) to using other models, such
as power relationships, to define a hierarchy. In contrast, the
Shipibo do not use a bifurcation to produce sex markings; rather
the male/female distinction is fundamental at the level of the
generation of a cognitive structure. Does this have a different
consequence for how female/male relationships are then emically modeled?

D. Read