It probably goes without saying that gender is an important category

It probably goes without saying that gender is an important category, not
only in real life, for real speakers, but also in the life of languages and in
their shape and structure. Most simply put, gender as a linguistic category
can be thought of as a means by which speakers, in their language, reflect
groupings of real-world elements along various dimensions. Admittedly,
a more nuanced view has to recognize that such groupings can be arbitrary
and lack (full) semantic motivation, but even in such cases it is usually
assumed that there once was a semantic basis to the groupings. Inasmuch
as these elements are usually the kinds of entities that in most languages
are given linguistic substance as nouns, it is typically in the nominal system
that gender becomes an issue, though once again, a more careful crosslinguistic
consideration reveals that gender distinctions can play a role in all parts of grammar. Just what these aforementioned dimensions are, of
course, varies considerably from language to language; some may have
semantic value, some merely grammatical value, but, importantly, all of
them give linguists a sense of glimpsing into what speakers deem – or
have deemed – to be salient for dividing up the totality of the physical,
emotional, spiritual, and even ethereal world.
One can think in terms of several methodologies that can inform the
study of gender: experimental, via carefully designed controlled elicitation
and experimentation on how speakers create and categorize novel objects;
typological, via surveying the range of types of classifications found in different
languages; analytic, via determination of the actual morphological
categories and marking, of the generalizations that speakers have made,
and of the semantics of particular classes (where appropriate); sociolinguistic,
via the investigation of the social value assigned to particular
class markers; and historical, via the examination of synchronic variation
(under the assumption that variation can signal change in progress), via
careful philological analysis of different stages of a language, via corpusbased
work (which of course can be synchronic in nature), via the study of
loanwords and how they are assigned to gender classes, or via reconstruction
and a consideration of what it tells us about unattested stages. Taken
together, these various methodologies add up to the rather remarkable fact
that gender is one of the few linguistic features to have spurred research
with such a broad range of methods, carried out by linguists of virtually all
theoretical persuasions.
Many of these methods are