instance on what basis one divides a particular people into different languages. Even the
criteria themselves are not perfect. Mass media and education may overcome barriers to
communication that would otherwise lead to lack of intelligibility on account of structural
linguistic differences. Alternatively, people often find reason to regard others as different,
even when their speech is mutually intelligible. Consequently, the enumeration of
languages and their speakers is fraught with difficulty, and needs to be treated with
caution.
1.2. Linguistic structure and relatedness
The subject matter of linguistics concerns the variety and nature of human languages,
their inner workings, structure and histories, and what those reveal about the nature of the
humanity, socially, cognitively and biologically. For these reasons, linguists generally
approach the identification of languages in taxonomic terms, by identifying language
families. Hence, we group Catalan, French, Italian, Spanish, Portuguese, Romanian and
Romansh together as Romance languages, recognizing their shared linguistic structure
and common origin in Latin, the language of the Roman Empire that dominated Europe
for several centuries. The Romance languages form one sub-family of the Indo-European
family, which embraces Celtic, Germanic, Indo-Aryan and Slavic, among others, as
additional sub-families.
Each recognized grouping indicates a degree of shared structure and a common
historical origin. Their relationship needs to be constructed from historical records, where
available, and by careful comparison of word forms and other structural properties of
language (the comparative method). Many times, written historical records are not
available, in which cases we must rely on archaeological evidence to assist in dating the
events that resulted in the current diversity of the family. In these circumstances, it is the
most recently developed language families that have left the most evidence of their
common origin, and which are most readily identified. Such families include Afro-
Asiatic (primarily found in Northern Africa and Western Asia), Austro-Asiatic
(Southeastern Asia), Indo-European (principally Europe, South and Central Asia, but
now spread throughout the globe), and Niger-Congo (sub-Saharan Africa), among other,
smaller families. There are generally acknowledged to be between as few as 16 (Comrie
1987) and as many as 108 (Gordon 1995) such family groups of languages still spoken in
the world.
Not all languages can be easily classified this way, and there are many isolated
languages as well as languages of indeterminate status. For example, Japanese and
Korean, in spite of superficial similarities, are rather different from each other and from
other languages of Eastern Asia. Despite efforts to connect them with the Altaic family
(including the Mongolian and Turkic languages of Central Asia), the Dravidian family
(including languages of Southern India and the isolate Brahui from Pakistan), no
proposed family affiliation for them has been widely accepted. Similarly, the languages
of Papua New Guinea, the Andaman Islands, and the isolate Kusunda (spoken in Nepal,
now possibly extinct) have been suggested to be related, but many linguists do not even
accept a family relationship among the languages of the Andaman Islands or within
Papua New Guinea, and prefer to recognize several distinct families within those
geographic groups. Generally these situations point to very old language communities,
sometimes going back to the earliest known prehistoric expansions of humanity into such
areas (Diamond 1997, 2005; Nichols 1992, 1998; Renfrew 1998). The communities in
question may presently subsist (or have subsisted in the known past) on hunting and
gathering using essentially Neolithic technologies. The wide dispersal of such languages,
and their small numbers of speakers attests to the large extent of their original domain,
and to their subsequent envelopment by newer and larger groups more recently.
The distribution of languages and language families around the world thus tells an
important story about the successive waves of human expansion throughout the habitable
areas of the globe. The relatively recent global expansion of the historically European
languages English, French, Portuguese, Russian and Spanish, under European
colonization and North American economic, political and cultural influence, is only the
most recent chapter in this story. Earlier expansions include the expansion of the Indo-
Europeans beginning about 6000 BC from a homeland possibly in Western Asia into the
South Asian subcontinent in the East, and most of Europe in the West; the Austronesian
expansion from Southern China in about 3500 BC throughout Oceania and later to Easter
Island, Hawaii, New Zealand and Madagascar; the Niger-Congo expansion from the
Sahel throughout most of sub-Saharan Africa; and similar expansions of Amerindian
languages in various historical phases from Northern through Southern America.
1.3. Language death and disappearance
An equally important part of this story is the extinction of large numbers of speech
varieties that existed before each such expansion. The most dramatic examples of
language extinctions in recent history occurred in the Americas and Australia. At the time
of Columbus, an unknown number of distinct languages, easily in the thousands, were
spoken in the Americas. Today, as a result of wars, disease, and incorporation into the
populations of European colonists, only a few hundred remain, and many of those
remaining, especially in North America, are near extinction or in danger of being
replaced by a European Language (Adelaar 1991, 2004; Cuaron and Lastra 1991; Dixon
and Aikhenvald 1999; Kinkade 1991; Krauss 1992; Mithuun 1999; Zepeda and Hill
1991). In Australia, out of more than 200 languages at the time of European arrival, about
50 languages have died out in the last 100 years, and 130 more have very few speakers
and are unlikely to survive much longer (Dixon 1991; Walsh 1991). In some places,
languages have disappeared so long ago or so completely that little is known of them.
Such is the case with the languages originally spoken in Tasmania for as much as 40,000
years, whose speakers all died, from warfare or disease, before much of their linguistic
heritage could be recorded. Likewise, none of the remaining Pygmy groups of sub-
Saharan Africa speak languages that appear to go back to their early occupation of the
region. Instead, they speak languages brought in later from the Niger-Congo family
originating in West Africa (Diamond 1997). Nothing is known today of the languages
they might have spoken prior to that time.