Introducing English Linguistics

Historical/archeological information.
As the previous two sections
have demonstrated, the comparative method relies quite heavily on lin-
guistic evidence to establish genetic relationships among languages.
However, non-linguistic evidence, such as historical information and
archeological evidence, can supplement linguistic evidence to help in the
classification of languages, especially to help date the origins of proto-lan-
guages for which no linguistic evidence exists.
The farther back in time one goes, the more sketchy historical information
about languages and their speakers becomes. This explains why we know so
little about either Proto-Indo-European or Proto-Germanic. In the case of
Proto-Indo-European, while the reconstruction of this language has, as Olson
(2003: 142) comments, provided considerable information concerning how
speakers of PIE lived, we currently have no hard evidence about “when and
where these people lived.” For this reason, we can only guess when this lan-
guage might have initially been spoken, who spoke it, and how migrations
of PIE speakers led to the development of sub-families of PIE (e.g. Proto-
Germanic). Dixon (1997: 48) states that although the common consensus is
that PIE began around 6,000 years ago, he notes that others have provided
evidence that the language could have originated up to 10,500 years ago. 
We can also only speculate about where PIE was initially spoken. The
most widely accepted view of the origins of PIE is the Kurgan Hypothesis,
which was originally proposed by the archeologist Marija Gimbutas
(1956). This hypothesis places the original speakers of PIE just north of the
Black Sea c. 6,000 years ago. Through a series of migrations, these speak-
ers spread their language all the way to Europe, spawning over time the
various sibling languages of PIE, including Proto-Germanic. Archeological
and linguistic evidence suggests that original speakers of PIE were war-
riors who rode horses as they made their way to Europe. An alternative
but much less widely accepted hypothesis is Renfrew’s (1987) farming-
The development of English
29

---

dispersal hypothesis. Contrary to Gimbutas, Renfrew argued that the orig-
inal speakers of PIE were not warriors but farmers, and that the spread of
farming from Anatolia (Turkey) to Greece and eventually Europe was
responsible for the spread of PIE. This hypothesis leads to a much earlier
dating of the origins of PIE to c. 10,500 years ago. One of the problems with
this hypothesis for historical linguists, Renfrew (2000: 14) acknowledges,
is that “they assume some specific chronological threshold beyond which
the techniques of the comparative method cannot penetrate.” In other
words, Renfrew’s dates for the origins of PIE extend beyond those for
which linguistic reconstruction can be reliably conducted and point to
the limitations of the comparative method.
The comparative method has clearly yielded valuable information
about languages and the extent to which they are related or unrelated.
However, this method has limitations, particularly with respect to how far
one can go back in time in the process of reconstructing languages and
language families. We have clear evidence that the Germanic branch of
Indo-European existed, and by examining languages grouped within the
language family, we can infer the existence of Proto-Indo-European. But
some linguists have attempted to go back further in time in the search for
ancestral languages to find, for instance, a larger super-family that would
include Proto-Indo-European. This process involves reconstructing a proto-
language on the basis of other proto-languages that in turn may them-
selves have been reconstructed from proto-languages. While some lin-
guists have argued that such a process is reliable, others have claimed that
vocabulary, for instance, changes so quickly that this endless process of
reconstruction is fraught with problems.
Greenberg (2000) has proposed a language family called Eurasiatic, which
includes language families such as Indo-European, Uralic, and Altaic as well
as other languages, such as Japanese and Korean, which have defied easy
classification into the major existent language families. Eurasiatic dates
back to c. 15,000 years ago and was reconstructed using a method called
mass lexical comparison. This method involves comparing sound similari-
ties between a set of common words in hundreds of languages. Statistical
tests are then conducted to determine the statistical probability that the
languages being compared are related. Another earlier language family that
has been proposed is called Nostratic. Some view this family as an alterna-
tive to Eurasiatic, others as a family that would include Eurasiatic. Still oth-
ers believe in the notion of monogenesis: the idea, as Trask (1996: 391)
observes, “that human language evolved only once, and that all languages
that have ever been spoken are descended from that single ancestor.” This
original language has been called Proto-World.
The problem with positing language families such as Eurasiatic and
Nostratic is that the reliability of one’s reconstruction diminishes consid-
erably if a hypothetical language family is reconstructed from other lan-
guage families. Moreover, if such reconstructions are based on compar-
isons of vocabulary, it is crucial that these comparisons be based on cog-
nate words, not borrowings. And in many cases it is difficult to determine
whether a given word in a language is a cognate or a borrowing. For these
30
INTRODUCING ENGLISH LINGUISTICS

---

reasons, many linguists remain skeptical of reconstructed language fami-
lies such as Eurasiatic and Nostratic.
While it may be difficult to precisely date the origins of PIE and Proto-
Germanic, we can be much more confident about the external history of
English. And knowledge of this history can be combined with surviving
linguistic evidence to provide a fairly precise description of the history of
the English language, and its various stages of development.