You are on page 1of 18

English, as we know it, developed in Britain and more recently in America and elsewhere

around the world. It did not begin in Britain but was an immigrant language,
coming there with the invading Anglo-Saxons in the ith century. Beore that,
English was spoken on the !ontinent, bordering on the "orth Sea. And even longer
beore, it had developed rom a speechway we call Indo-European, which was the
source o most other European and many south-Asian languages. #e have no historical
records o that prehistoric tongue, but we know something about it and the
people who spoke it rom the comparisons linguists have made between the various
languages that eventually developed rom it.
Indo-European is a matter o culture, not o genes. $he contrast between our
genetic inheritance and the language we speak is highlighted by some recent discoveries
in genetics. Scholars used to think o early Europe as inhabited by a %aleolithic
&old Stone Age' people who were hunter-gatherers but whose culture was replaced
by "eolithic &new Stone Age' agriculturalists. $he latter were supposedly replaced
by a Bron(e Age culture &beginning between )*** and +*** B.!.', spread by a
sweeping invasion o technologically more advanced people rom the east.
,ecent genetic studies, however, have established that most modern Europeans
&and o course the Americans descended rom them' owe only about -* percent o
their biological inheritance to the later peoples and .* percent to their early %aleolithic
ancestors. It looks now as though the genetic characteristics o Europeans have been
remarkably stable, despite the striking changes that have overtaken European culture
between earliest times and the beginning o recorded history.
/inguists have also long thought that the Indo-European languages, o which
English is one, were spread across the !ontinent by the invading Bron(e Age
hordes, who came in chariots and wiped out the native populations and cultures.
0ore recently, however, it has been posited that Indo-European languages were
spread throughout Europe very much earlier, and that the Indo-European expansion
did not ollow a simple east-to-west path, but was ar more complex and
included a south-to-north migration o early !eltic and 1ermanic peoples rom
Spain and southern 2rance. At the present time all that can be said conidently about the
early history o the Indo-European languages is that we know less than
we ormerly thought we did. 3et we do know some things.
I"45-E6,5%EA" 5,I1I"S
Indo-European !ulture
5n the basis o cognate words, we can iner a good deal about Indo-European culture
beore it spread over many parts o Europe and Asia. $hat spread started
no later than the third or ourth millennium B.!. and perhaps very much earlier.
Indo-European culture was considerably advanced. $hose who spoke the parent
language, which we call %roto-Indo-European, had a complex system o amily
relationships.
$hey could count. $hey used gold and perhaps silver also, but copper
and iron only later. $hey drank a honey-based alcoholic beverage whose name has
come down to us as mead. #ords corresponding to wheel, axle, and yoke make it
clear that they used wheeled vehicles. $hey were small armers, not nomads, who
worked their ields with plows, and they had domesticated animals and owl.
$heir religion was polytheistic, including a Sky 2ather &whose name is preserved
in the ancient 7edic hymns o India as 4yaus pitar, in 1reek myth as 8eus
pat9r, among the ,omans as :upiter, and among the 1ermanic peoples as $iw, or
whom $uesday is named'. $he cow and the horse were important to their society,
wealth being measured by a count o cattle; the /atin word pecus meant <cattle= but
was the source o the word pecnia <wealth,= rom which we get pecuniary> and our
word ee comes rom a related 5ld English word oh, which also meant both
<cattle= and <wealth.= So we know things about the ancient Indo-European speakers
on the basis o orms that were not actually recorded until long ater Indo-
European had ceased to be a single language.
$he Indo-European ?omeland
#e can only guess where Indo-European was originally spoken@but there are
clues, such as plant and animal names. !ognate terms or trees that grow in temperate
climates &alder, apple, ash, aspen, beech, birch, elm, ha(el, linden, oak, willow,
yew', coupled with the absence o such terms or 0editerranean or Asiatic trees
&olive, cypress, palm'> cognate terms or wol, bear, lox &5ld English leax <salmon=',
but none or creatures indigenous to Asia@all this points to an area between northern
Europe and southern ,ussia as the home o Indo-European beore its dispersion.
And the absence o a common word or ocean suggests, though it does not
in itsel prove, that this homeland was inland.
$he early Indo-Europeans have been identiied with the Aurgan culture o
mound builders who lived northwest o the !aucasus and north o the !aspian Sea
as early as the ith millennium B.!. &1imbutas, Aurgan !ulture'. $hey domesticated
cattle and horses, which they kept or milk and meat as well as or transportation.
$hey combined arming with herding and were a mobile people, using our-wheeled
wagons to cart their belongings on their treks. $hey built ortiied palaces on hilltops
&we have the Indo-European word or such orts in the polis o place names like
Indianapolis and in our word police', as well as small villages nearby. $heir society was a
stratiied one, with a warrior nobility and a common laboring class. In addition
to the sky god associated with thunder, the sun, the horse, the boar, and the
snake were important in their religion. $hey had a highly developed belie in lie
ater death, which led them to the construction o elaborate burial sites, by which
their culture can be traced over much o Europe. Early in their history, they
expanded into the Balkans and northern Europe, and thereater into Iran, Anatolia,
and southern Europe.
5ther locations have also been proposed or the Indo-European homeland,
such as north-central Europe between the 7istula and the Elbe and eastern
Anatolia &modern $urkey and the site o the ancient ?ittite empire'. $he dispersal
o Indo-European was so early that we may never be sure o where it began or o
the paths it ollowed.
?ow Indo-European #as 4iscovered
Even a casual comparison o English with some other languages reveals similarities
among them. $hus English ather clearly resembles "orwegian, 4anish, and
Swedish ader, Icelandic aBir, 4utch vader, and 1erman 7ater &especially when
one is aware that the letter v in 1erman represents the same sound as '.
Although there is still a air resemblance, the English word is not Cuite so similar
to /atin pater, Spanish padre, %ortuguese pai, !atalan pare, and 2rench pDre.
1reek patr, Sanskrit pitar-, and %ersian pedar are all strikingly like the /atin
orm, and &allowing or the loss o the irst consonant' 1aelic athair resembles the
others as well. It takes no great insight to recogni(e that those words or <ather=
are somehow the Esame.F Because such similarity o words is reinorced by other
parallels among the languages, we are orced to look or some explanation o the
resemblances.
$he explanation@that all those languages are historical developments o a no
longer existing source language@was irst proposed several centuries ago by Sir
#illiam :ones, a British Gudge and Sanskrit scholar in India. $he Indo-European
hypothesis, as it is called, is now well supported with evidence rom many languages;
a language once existed that developed in dierent ways in the various
parts o the world to which its speakers traveled. #e call it %roto-Indo-European
&or simply Indo-European' because at the beginning o historical times languages
derived rom it were spoken rom Europe in the west to India in the east. Its
Edescendants,F
which make up the Indo-European amily, include all o the languages
mentioned in the preceding paragraph, as well as ,ussian, %olish, !(ech, Bulgarian,
Albanian, Armenian, ,omany, and many others.
"ineteenth-century philologists sometimes called the Indo-European amily o
languages Aryan, a Sanskrit term meaning <noble,= which is what some o the languages=
speakers immodestly called themselves. Aryan has also been used to name
the branch o Indo-European spoken in Iran and India, now usually reerred to as
Indo-Iranian. $he term Aryan was, however, generally given up by linguists ater
the "a(is appropriated it or their supposedly master race o "ordic eatures, but
it is still ound in its original senses in some older works on language. $he term
Indo-European has no racial connotations> it reers only to the culture o a group
o people who lived in a relatively small area in early times and who spoke a more or less
uniied language out o which many languages have developed over thousands
o years. $hese languages are spoken today by approximately hal o the
world=s population.
/A"16A1E $3%5/513 A"4 /A"16A1E 2A0I/IES
In talking about a language amily, we use metaphors like EmotherF and EdaughterF
languages and speak o degrees o Erelationship,F Gust as though languages had
ospring that could be plotted on a genealogical, or amily-tree, chart. $he terms
are convenient ones> but, in the discussion o linguistic EamiliesF that ollows, we
must bear in mind that a language is not born, nor does it put out branches like a
tree@nor, or that matter, does it die, except when every single one o its speakers
dies, as has happened to Etruscan, 1othic, !ornish, and a good many other languages.
#e speak o /atin as a dead language, but in act it still lives in Italian,
2rench, Spanish, and the other ,omance languages. In the same way, %roto-
Indo-European continues in the various present-day Indo-European languages,
including English.
?ence the terms amily, ancestor, parent, and other genealogical expressions
applied to languages are metaphors, not literal descriptions. /anguages are
developments
o older languages rather than descendants in the sense in which people are descendants
o their ancestors. $hus Italian and Spanish are dierent developments o an
earlier, more uniied /atin. /atin, in turn, is one o a number o developments o a still
earlier language called Italic. Italic, in its turn, is a development o Indo-European.
Earlier scholars classiied languages as isolating, agglutinative, incorporative,
and inlective, exempliied respectively by !hinese, $urkish, Eskimo, and /atin.
$he isolating languages were once thought to be the most primitive type; they
were languages in which each idea was expressed by a separate word and in
which the words tended to be monosyllabic. But although !hinese is an isolating
and monosyllabic language in its modern orm, its earliest records &rom the middle
o the second millennium B.!.' represent not a primitive language but actually one
in a late stage o development. 5ur prehistoric ancestors did not prattle in onesyllable
words.
Earlier scholars also observed, Cuite correctly, that in certain languages, such as
$urkish and ?ungarian, words were made up o parts Estuck together,F as it were>
hence the term agglutinative &etymologically <glued to='. In such languages the elements
that are put together are usually whole syllables having clear meanings. $he inlectional
suixes o the Indo-European languages were supposed once to have been independent
words> hence some early scholars believed that the inlective languages had
grown out o the agglutinative. /ittle was known o what were called incorporative
languages, in which maGor sentence elements are combined into a single word.
$he trouble with such a classiication is that it was based on the now discarded
theory that early peoples spoke in monosyllables. 2urthermore, the dierence
between agglutinative and inlective languages was not well deined, and there was
considerable overlapping. "evertheless, the terms are widely used in the description
o languages. 5bGective and well-inormed typological classiication has been especially
useul in showing language similarities and dierences &1reenberg, /anguage
$ypology'. 2rom the historical point o view, however, much more satisactory is the
genetic classiication o languages, made on the basis o such correspondences o
sound and structure as indicate relationship through common origin. %erhaps the
greatest contribution o nineteenth-century linguistic scholars was the painstaking
investigation o those correspondences, many o which had been casually noted
long beore.
"5"-I"45-E6,5%EA" /A"16A1ES
Beore proceeding to a more detailed discussion o the Indo-European group,
we look briely at those languages and groups o languages that are not Indo-
European. $wo important groups have names that relect the biblical attempt to
derive all human races rom the three sons o "oah; the Semitic &rom the /atin
orm o the name o his eldest son, more correctly called Shem in English' and the
?amitic &rom the name o his second son, ?am'. $he term :aphetic &rom "oah=s
third son, :apheth', once used or Indo-European, has long been obsolete. 5n the
basis o many phonological and morphological eatures that they share, Semitic
and ?amitic are thought by many scholars to be related through a hypothetical
common ancestor, ?amito-Semitic, or Aroasiatic, as it is usually called now.
$he Semitic group includes the ollowing languages in three geographical subgroups;
&Eastern' Akkadian, whose varieties include Assyrian and Babylonian>
&#estern' ?ebrew, Aramaic &the native speech o :esus !hrist', %hoenician, and
0oabitic> and &Southern' Arabic and Ethiopic. 5 these, only Arabic is spoken by
large numbers o people over a widespread area. ?ebrew has been revived comparatively
recently in Israel, to some extent or nationalistic reasons. It is interesting to
note that two o the world=s most important religious documents are written in
Semitic languages@the :ewish scriptures or 5ld $estament in ?ebrew &with large
portions o the books o E(ra and 4aniel in Aramaic' and the Aoran in Arabic.
$o the ?amitic group belong Egyptian &called !optic ater the close o the third
century o the !hristian era', the Berber dialects o "orth Arica, various !ushitic
dialects spoken along the upper "ile &named or !ush, a son o ?am', and !hadic
in !had and "igeria. Arabic became dominant in Egypt during the sixteenth century,
when it replaced !optic as the national language.
?amitic is unrelated to the other languages spoken in central and southern
Arica, the vast region south o the Sahara. $hose sub-Saharan languages are usually
classiied into three main groups; "ilo-Saharan, extending to the eCuator, a large
and highly diversiied group o languages whose relationships with one another are
uncertain> "iger-Aordoanian, extending rom the eCuator to the extreme south, a
large group o languages o which the most important belong to the Bantu group,
including Swahili> and the Ahoisan languages, such as ?ottentot and Bushman, spoken
by small groups o people in the extreme southwestern part o Arica. 7arious
o the Ahoisan languages use clicks@the kind o sound used by English speakers as
exclamations and conventionally represented by spellings such as tsk-tsk and cluckcluck,
but used as regular speech sounds in Ahoisan and transcribed by slashes or
exclamation points, as in the H5Hkung language, spoken in Angola.
In south Asia, languages belonging to the 4ravidian group were once spoken
throughout India, where the earlier linguistic situation was radically aected by the Indo-
European invasion o approximately IJ** B.!. $hey are the aboriginal languages
o India but are now spoken mainly in southern India, such as $amil and $elegu.
$he Sino-$ibetan group includes the various languages o !hina, such as
!antonese and 0andarin, as well as $ibetan, Burmese, and others. :apanese
is unrelated to !hinese, although it has borrowed the !hinese written characters
and many !hinese words. It and Aorean are sometimes thought to be members o
the Altaic amily, mentioned below, but the relationship is not certain. Ainu, the
language o the aborigines o :apan, is not clearly related to any other language.
A striking characteristic o the Austronesian &or 0alayo-%olynesian' languages
is their wide geographical distribution in the islands o the Indian and the %aciic
oceans, stretching rom 0adagascar to Easter Island. $hey include 0alay, 0aori
in "ew 8ealand, ?awaiian, and other %olynesian languages. $he native languages
o Australia, spoken by only a ew aborigines there nowadays, have no connection
with Austronesian, nor have the more than a hundred languages spoken in "ew
1uinea and neighboring islands.
American Indian languages are a geographic rather than a linguistic grouping,
comprising many dierent language groups and even isolated languages having little
or no relationship with one another. A very important and widespread group o
American Indian languages is known as the 6to-A(tecan, which includes "ahuatl,
the language spoken by the A(tecs, and various closely related dialects. Aleut and
Eskimo, which are very similar to each other, are spoken in the Aleutians and all
along the extreme northern coast o America and north to 1reenland. In the Andes
0ountains o South America, Aechumaran is a language stock that includes Aymara
and Kuechua, the speech o the Incan Empire. $he isolation o the various groups,
small in number to begin with and spread over so large a territory, may account to
some extent or the great diversity o American Indian tongues.
BasCue, spoken in many dialects by no more than hal a million people in the
region o the %yrenees, has always been something o a popular linguistic mystery.
It now seems airly certain, on the basis o coins and scanty inscriptions o the
ancient Iberians, that BasCue is related to the almost completely lost language o
those people who once inhabited the Iberian peninsula and in "eolithic times were
spread over an even larger part o Europe.
An important group o non-Indo-European languages spoken in Europe, as well
as in parts o Asia, is the 6ral-Altaic, with its two subgroups; the 6ralic and the
Altaic. 6ralic has two branches; Samoyed, spoken rom northern European ,ussia
into Siberia, and 2inno-6gric, including 2innish, Estonian, /appish, and ?ungarian.
Altaic includes several varieties o $urkish, such as 5ttoman $urkish &5smanli' and
the languages o $urkestan and A(erbaiGan, as well as 0ongolian and 0anchu.
$he oregoing is by no means a complete survey o non-Indo-European languages.
It includes only some o the most important groups and individual languages.
In A 1uide to the #orld=s /anguages, 0erritt ,uhlen lists IL phyla &large groups o
distantly related languages', including nearly +** maGor groups and subgroups and
about J*** languages, o which I)* are Indo-European. Although Indo-European
languages are ewer than + percent o the number o languages in the world, nearly
hal the world=s population speaks them.
/anguages may be related to each other more distantly in superamilies.
:oseph 1reenberg has posited a linguistic stock called Eurasiatic, which includes Indo-
European, 6ral-Altaic, and other languages such as Etruscan, Aorean,
:apanese, Aleut, and Eskimo. 5ther linguists have posited even larger superamilies,
such as "ostratic, which includes many languages o Europe, Asia, Arica, and
"orth America. 5thers ask whether all human languages can be traced to a single
original speech, %roto-#orld or %roto-?uman. But no one knows> we are Cuite in
the dark about how it all began.
0AI" 4I7ISI5"S 52 $?E I"45-E6,5%EA" 1,56%
Some Indo-European languages@or example, $hracian, %hrygian, 0acedonian,
and Illyrian@survive only in scanty remains. It is likely that others have disappeared
without leaving any trace. 0embers o the ollowing subgroups survive as
living tongues; Indo-Iranian, Balto-Slavic, ?ellenic, Italic, !eltic, and 1ermanic.
Albanian and Armenian are also Indo-European but do not it into any o these
subgroups. Anatolian and $ocharian are no longer spoken in any orm.
$he Indo-European languages are either satem languages or centum languages.
Satem and centum are respectively the Avestan &an ancient Iranian language' and
/atin words or hundred. $he two groups are dierentiated by their development o
Indo-European palatal k.
In Indo-European, palatal k &as in Mkmtom <hundred=' was a distinct phoneme
rom velar k &as in the verbal root Mkwer- <do, make,= which we have in the Sanskrit
loanword karma and in the name Sanskrit itsel, which means something like <wellmade='.
&An asterisk beore a orm indicates that it is a reconstruction based on comparative
study.' In the satem languages@Indo-Iranian, Balto-Slavic, Armenian, and
Albanian@the two k sounds remained separate phonemes, and the palatal k became
a sibilant@or example, Sanskrit &Indic' atam, /ithuanian &Baltic' Nimtas, and 5ld
!hurch Slavic sto. In the other Indo-European languages, the two k sounds became
a single phoneme, either remaining a k, as in 1reek &?ellenic' &he'katon and #elsh
&!eltic' cant, or shiting to h in the 1ermanic group, as in 5ld English hund &our hundred
being a compound in which -red is a development o an originally independent
word meaning <number='. In general, the centum languages tend to be spoken in the
#est and the satem languages in the East, although $ocharian, the easternmost o all
Indo-European tongues, belongs to the centum group.
Indo-Iranian
$he Indo-Iranian group &Iranian is rom the same root as the word Aryan' is one o
the oldest or which we have historical records. $he 7edic hymns, written in an
early orm o Sanskrit, date rom at least I*** B.!. but relect a poetic tradition
stretching back to the second millennium B.!. !lassical Sanskrit appears about
J** B.!. It is much more systemati(ed than 7edic Sanskrit, or it had been sei(ed
upon by early grammarians who ormulated rules or its proper use> the very
name Sanskrit means <well-made= or <perected.=
$he most remarkable o the Indian grammarians was %anini. About the same
time &ourth century B.!.' that the 1reeks were indulging in anciul speculations
about language and in antastic etymologi(ing, he wrote a grammar o Sanskrit that to
this day holds the admiration o linguistic scholars. 5ther ancient Indian
scholars also wrote works preserving the language o the old sacred literature that
put much o the grammatical writing o the 1reeks and ,omans to shame. Sanskrit
is still written by Indian scholars according to the old grammarians= rules. It is in
no sense dead as a written language but has a status much like that o /atin in
medieval and ,enaissance Europe.
Indic dialects had developed long beore Sanskrit became a reined and learned
language. $hey are called %rakrits &a name that means <natural,= contrasting with
the Ewell-made-nessF o Sanskrit', and some o them@notably %ali, the religious
language o Buddhism@achieved high literary status. 2rom these %rakrits are indirectly
derived the various non-4ravidian languages o India, the most widely
known o which are Bengali, ?indi, and 6rdu.
,omany &1ypsy' is also an Indic dialect, with many loanwords rom other languages
acCuired in the course o the ,omanies= wanderings. #hen they irst appeared
in Europe in the late 0iddle Ages, many people supposed them to be Egyptians@
whence the name 1ypsy. A long time passed beore the study o their language
revealed that they had come originally rom northwestern India. $he name ,omany
has nothing to do with ,ome, but is derived rom the word rom <human being.=
/ikewise the rye o ,omany rye &that is, <,omany gentleman=' has nothing to do
with the cereal crop, but is a word akin to Sanskrit raGan <king,= as well as to /atin
rex, 1erman ,eich, and English regal and royal &rom /atin and 2rench'.
$hose Indo-Europeans who settled in the Iranian %lateau developed several languages.
5ld %ersian is the ancestor o modern Iranian. It was the language o the
district known to the 1reeks as %ersis, whose inhabitants under the leadership o
!yrus the 1reat in the sixth century B.!. became the predominant tribe. 0any
%ersians migrated to India, especially ater the 0uslim conCuest o Iran in the
eighth century. $hey were 8oroastrians in religion who became the ancestors
o the modern %arsis &that is, %ersians' o Bombay. Avestan, another Iranian
tongue, is a sacred language, preserved in the Avesta, a religious book ater which
the language is named. $here are no modern descendants o Avestan, which was
the language o the sage 8arathustra@8oroaster to the 1reeks.
Armenian and Albanian
Armenian and Albanian are independent subgroups. $he irst has in its word stock
so many %ersian loanwords that it was once supposed to belong to the Indo-Iranian
group> it also has many borrowings rom 1reek and rom Arabic and Syrian.
Albanian also has a mixed vocabulary, with words rom Italian, Slavic, $urkish,
and 1reek. It is possibly related to the ancient language o Illyria in an Illyrian
branch o Indo-European. Evidence o the ancient language is so meager, however,
and modern Albanian has been so much inluenced by neighboring languages that
it is diicult to tell much about its ainities.
$ocharian
$he $ocharian language has two varieties, called $ocharian A &an eastern dialect'
and $ocharian B &a western dialect'. $he language is misnamed. #hen it was discovered
at the end o the nineteenth century in some volumes o Buddhist scriptures
and monastic business accounts rom central Asia, it was at irst thought to be
a orm o Iranian and so was named ater an extinct Iranian people known to the
ancient 1reek geographer Strabo as $ocharoi. /ater it was discovered that $ocharian
is linguistically Cuite dierent rom Iranian. "evertheless, the name has stuck> the
language itsel has long been extinct.
Anatolian
Shortly ater the discovery o $ocharian, another group o Indo-European languages
was identiied in Asia 0inor. Excavations uncovered the royal archives at
the capital city o the ?ittites, a people mentioned in the 5ld $estament and in
Egyptian records rom the second millennium B.!. $hose archives included works
in a number o ancient languages, including one otherwise unknown. As the writings
in the unknown tongue were deciphered, it became clear that the language,
?ittite, was Indo-European, although it had been prooundly inluenced by non-
Indo-European languages spoken around it. /ater scholars identiied several dierent
but related languages &/uwian, %alaic, and /ydian', and the new branch was
named Anatolian, ater the area where it was spoken. 5ne o the interesting eatures
o ?ittite is that it preserves an Indo-European ElaryngealF sound &transliterated
h' that was lost in all o the other Indo-European languages &or example, in
?ittite pahhur <ire= compared with 1reek pOr, 6mbrian pir, !(ech pP, $ocharian
por, and 5ld English yQ r'.
Balto-Slavic
Although the oldest records o the Baltic and the Slavic languages show them as
Cuite dierent, most scholars have assumed a common ancestor closer than Indo-
European, called Balto-Slavic. $he chie Baltic language is /ithuanian, and the
closely related /atvian is spoken to its north. /ithuanian is Cuite conservative
phonologically, so that one can ind a number o words in it that are very similar
in orm to cognate words in older Indo-European languages@or example,
/ithuanian 4ivas and Sanskrit devas <god= or /ithuanian platRs and 1reek platSs
<broad.=
Still another Baltic language, 5ld %russian, was spoken as late as the seventeenth
century in what is now called East %russia. %russians, like /ithuanians and /atvians,
were heathens until the end o the 0iddle Ages, when they were converted at the
point o the sword by the Anights o the $eutonic 5rder@a military order that was
an outcome o the !rusades. $he aristocracy o the region &their descendants are the
%russian :unkers' came to be made up o members o this order, who, having saved
the souls o the heathen Balts, proceeded to take over their lands.
Slavic alls into three main subdivisions. East Slavic includes ,ussian, 6krainian,
and Belarussian, spoken in Belarus, north o the 6kraine. #est Slavic includes
%olish, !(ech, the similar Slovak, and Sorbian &or #endish', a language spoken
by a small group o people in eastern 1ermany. $he South Slavic languages include
Bulgarian, Serbo-!roatian, and Slovenian. $he oldest Slavic writing we know is in 5ld
!hurch Slavic &or Slavonic', which remained a liturgical language long ater it
ceased to be generally spoken.
?ellenic
In ancient times there were many ?ellenic dialects, among them 0ycenaean, Aeolic,
4oric, and Attic-Ionic. Athens came to assume tremendous prestige, so its dialect,
Attic, became the basis o a standard or the entire 1reek world, a koine or <common
TdialectU,= which was ultimately to dominate the other ?ellenic dialects. 0ost
o the local dialects spoken in 1reece today, as well as the standard language, are
derived rom Attic. 4espite all their glorious ancient literature, the 1reeks have not
had a modern literary language until comparatively recently. $he new literary standard
makes considerable use o words revived rom ancient 1reek, as well as a
number o ancient inlectional orms> it has become the ordinary language o the
upper classes. Another development o the Attic koine, spoken by the masses, is
called demotike <popular.=
Italic
In ancient Italy, the main Indo-European language was /atin, the speech o /atium,
whose chie city was ,ome. 5scan and 6mbrian have long been thought to be sister
languages o /atin within the Italic subamily, but they may be members o an
independent branch o Indo-European whose resemblance to /atin is due to the
long period o contact between their speakers. It is well known that languages,
even unrelated ones, that are spoken in the same area and share bilingual speakers
&in an association called a Sprachbund' will inluence one another and thus become
more alike.
/atin became the most important language o the peninsula. As ,ome came
to dominate the 0editerranean world, it spread its inluence into 1aul, Spain, and
the Illyrian and 4anubian countries &and even into Britain, where /atin ailed to
displace !eltic'. $hus its language became a koine, as the dialect o Athens had
been earlier. Spoken /atin survives in the ,omance languages. It was Cuite dierent
rom the more or less artiicial literary language o !icero. All the ,omance languages@
such as Italian, Spanish, !atalan, 1alician, %ortuguese, 2rench, %rovenVal,
and ,omanian@are developments o 7ulgar /atin &so called because it was the
speech o the vulgus <common people=' spoken in various parts o the late ,oman
Empire.
2rench dialects have included "orman, the source o the Anglo-"orman dialect
spoken in England ater the "orman !onCuest> %icard> and the dialect o
%aris and the surrounding regions &the Wle-de-2rance', which or obvious reasons
became standard 2rench. In southern Belgium a dialect o 2rench, called
#alloon, is spoken. $he varieties o 2rench spoken in Kuebec, "ova Scotia,
"ew Brunswick, and /ouisiana are all developments o the dialects o northern
2rance and are no more EcorruptionsF o standard &0odern' 2rench than
American English is o present standard British. $he !aGuns &that is, Acadians' o
/ouisiana are descendants o exiles rom "ova Scotia, which was earlier a 2rench
colony called Acadia.
$he speech o the old kingdom o !astile, the largest and central part o Spain,
became standard Spanish. $he act that Spanish America was settled largely by people
rom southern Andalusia rather than rom !astile accounts or the most important
dierences in pronunciation between /atin American Spanish and the standard
language o Spain.
Because o the cultural preeminence o $uscany during the Italian ,enaissance,
the speech o that region@and speciically o the city o 2lorence@became standard
Italian. Both 4ante and %etrarch wrote in this orm o Italian. ,haeto-,omanic
comprises a number o dialects spoken in the most easterly Swiss canton and in the
$yrol.
!eltic
!eltic shows such striking correspondences with Italic in its verbal system and
inlectional endings that the relationship between them must have been close,
though not so close as that between Indic and Iranian or Baltic and Slavic. Some
scholars thereore group them together as developments o a branch they call
Italo-!eltic.
$he !elts were spread over a huge territory in Europe long beore the emergence
in history o the 1ermanic peoples. Beore the beginning o the !hristian
era, !eltic languages were spoken over the greater part o central and western
Europe. By the latter part o the third century B.!. !elts had spread even to Asia
0inor, in the region called or them 1alatia &part o modern $urkey', to whose
inhabitants Saint %aul wrote one o his epistles. $he !eltic language spoken in
1aul &1aulish' gave way completely to the /atin spoken by the ,oman conCuerors,
which was to develop into 2rench.
,oman rule did not prevent the British !elts rom using their own language,
although they borrowed a good many words rom /atin. But ater the Angles,
Saxons, and :utes arrived, British &Brythonic' !eltic was more severely threatened.
It survived, however, and produced a distinguished literature in the later 0iddle
Ages, including the 0abinogion and many Arthurian stories. In recent years, #elsh
&!ymric' has been actively promoted or nationalistic reasons. Breton is the language
o the descendants o those Britons who, at or beore the time o the Anglo-Saxon
invasion o their island, crossed the !hannel to the !ontinent, settled in the
1aulish province o Armorica, and named their new home or their old one@
Brittany. Breton is thus more closely related to #elsh than to long-extinct 1aulish.
$here have been no native speakers o !ornish, another Brythonic language, since
the early nineteenth century. Eorts have been made to revive it; church services
are sometimes conducted in !ornish, and the language is used in antiCuarian recreations
o the !eltic 0idsummer Eve rituals@but such eorts seem more sentimental
than practical.
It is not known whether %ictish, preserved in a ew glosses and place-name elements,
was a !eltic language. It was spoken by the %icts in the northwestern part o
Britain, where many 1aelic !elts also settled. $he latter were settlers rom Ireland called
Scots &Scotti', hence the name o their new home, Scotia or Scotland. $he
!eltic language that spread rom Ireland, called 1aelic or 1oidelic, was o a type
somewhat dierent rom that o the Britons. It survives in Scots 1aelic, sometimes
called Erse, a word that is simply a variant o Irish. 1aelic is spoken in the remoter
parts o the Scottish highlands and the 5uter ?ebrides and in "ova Scotia. In a
somewhat dierent development called 0anx, it survived until recently on the Isle
o 0an.
In Ireland, which was little aected by either the ,oman or the later Anglo-
Saxon invasions, Irish 1aelic was gradually replaced by English. It has survived in
some o the western counties, though most o its speakers are now bilingual. Eorts
have been made to revive the language or nationalistic reasons in Eire, and it is
taught in schools throughout the land> but this resuscitation, less successul than
that o ?ebrew in modern Israel, cannot be regarded as in any sense a natural
development.
In striking contrast to their wide distribution in earlier times, today the !eltic
languages are restricted to a ew relatively small areas abutting the Atlantic 5cean
on the northwest coast o Europe.
1ermanic
$he 1ermanic group is particularly important or us because it includes English.
5ver many centuries, certain radical developments occurred in the language spoken
by those Indo-European speakers living in 4enmark and the regions thereabout.
%roto-1ermanic &or simply 1ermanic', our term or that language, was relatively
uniied and distinctive in many o its sounds, inlections, accentual system, and
word stock.
6nortunately or us, those who spoke this particular development o Indo-
European did not write. %roto-1ermanic is to 1erman, 4utch, the Scandinavian
languages, and English as /atin is to Italian, 2rench, and Spanish. But %roto-
1ermanic, which was probably being spoken shortly beore the beginning o the
!hristian era, must be reconstructed Gust like Indo-European, whereas /atin is
amply recorded.
Because 1ermanic was spread over a large area, it eventually developed marked
dialectal dierences leading to a division into "orth 1ermanic, #est 1ermanic, and
East 1ermanic. $he "orth 1ermanic languages are 4anish, Swedish, "orwegian,
Icelandic, and 2aeroese &very similar to Icelandic and spoken in the 2aeroe Islands
o the "orth Atlantic between Iceland and 1reat Britain'.
$he #est 1ermanic languages are ?igh 1erman, /ow 1erman &%lattdeutsch',
4utch &and the practically identical 2lemish', 2risian, and English. 3iddish developed
rom medieval ?igh 1erman dialects, with many words rom ?ebrew and
Slavic. Beore #orld #ar II, it was a sort o international language o the :ews,
with a literature o high Cuality. Since that time, it has declined greatly in use,
with most :ews adopting the language o the country in which they live> and its
decline has been accelerated by the revival o ?ebrew in Israel. Arikaans is a
development
o seventeenth-century 4utch spoken in South Arica. %ennsylvania 4utch
&that is, 4eutsch' is actually a ?igh 1erman dialect spoken by descendants o early
American settlers rom southern 1ermany and Swit(erland. $he only East 1ermanic
language o which we have any detailed knowledge is
1othic. It is the earliest attested o all 1ermanic languages, aside rom a ew proper
names recorded by classical authors, a ew loanwords in 2innish, and some runic
inscriptions ound in Scandinavia. Almost all our knowledge o 1othic comes
rom a translation mainly o parts o the "ew $estament made in the ourth century
by #ulila, bishop o the 7isigoths, those 1oths who lived north o the
4anube ,iver. /ate as they are in comparison with the literary records o Sanskrit,
Iranian, 1reek, and /atin, these remains o 1othic provide us with a clear picture
o a 1ermanic language in an early stage o development and hence are o tremendous
importance to the history o 1ermanic languages.
1othic as a spoken tongue disappeared a long time ago without leaving a trace.
"o modern 1ermanic languages are derived rom it, nor do any o the other
1ermanic languages have any 1othic loanwords. 7andalic and Burgundian were
apparently also East 1ermanic in structure, but we know little more o them than
a ew proper names.
4uring the eighteenth-century EAge o ,eason,F the term 1othic was applied
to the Edark agesF o the medieval period as a term o contempt, and hence to the
architecture o that period to distinguish it rom classical building styles. $he general
eighteenth-century sense o the word was <barbarous, savage, in bad taste.=
/ater the term was used or the type onts ormerly used to print 1erman &also
called black letter'. $hen it denoted a genre o novel set in a desolate or remote
landscape, with mysterious or macabre characters and oten a violent plot. 0ore
recently it was applied to an outrX style o dress, cosmetics, and coiure, eaturing
the color black and accompanied by heavy metal adornments and body piercing in
unlikely parts o the anatomy. $hus the name o a people and a language long ago
lost to history survives in uses that have nothing to do with the 1oths and would
doubtless have both pu((led and ama(ed them.
!51"A$E #5,4S I" $?E I"45-E6,5%EA" /A"16A1ES
#ords that come rom the same source are said to be cognate &/atin co- and gnatus
<born together='. $hus the verb roots meaning <bear, carry= in Sanskrit &bhar-',
1reek &pher-', /atin &er-', 1othic &bair-', and 5ld English &ber-' are cognate, all
being developments o Indo-European Mbher-. !ognate words do not necessarily
look similar because their relationship may be disguised by sound changes that
have aected their orms dierently. $hus, English work and 1reek ergon are
supericially unlike, but they are both developments o Indo-European Mwergom
and thereore are cognates. Sometimes, however, there is similarity@or example,
between /atin ignis and Sanskrit agnis rom Indo-European Megnis <ire,= a root
that is unrelated to the other words or <ire= cited earlier, but that English has in
the /atin borrowing ignite.
Some cognate words have been preserved in many or even all Indo-European
languages. $hese common related words include the numerals rom one to ten, the
word meaning the sum o ten tens &cent-, sat-, hund-', words or certain bodily
parts &related, or example, to heart, lung, head, oot', words or certain natural
phenomena &related, or example, to air, night, star, snow, sun, moon, wind', certain
plant and animal names &related, or example, to beech, corn, wol, bear',
and certain cultural terms &related, or example, to yoke, mead, weave, sew'.
!ognates o practically all our taboo words@those monosyllables that pertain to
sex and excretion and that seem to cause great pain to many people@are to be
ound throughout the Indo-European languages. ?istorically, i not socially, those
ancient words are Gust as legitimate as any others.
It takes no special training to perceive the correspondences between the ollowing
words;
/atin 1reek #elsh English Icelandic 4utch
nus oinI un one einn een
duo duo dau two tveir twee
trs treis tri three YrZr drie
I<one-spot on a die=
!omparison o the orms or the number <two= indicates that non-1ermanic TdU
&as in the /atin, 1reek, and #elsh orms' corresponds to 1ermanic TtU &English,
Icelandic, and 4utch'. A similar comparison o the orms or the number <three=
indicates that non-1ermanic TtU corresponds to 1ermanic TU, the initial sound o
three and YrZr in English and Icelandic. Allowing or later changes@as in the case
o TU, which became TdU in 4utch, as also in 1erman &drei <three=', and TtU in
4anish, "orwegian, and Swedish &tre'@these same correspondences are perectly
regular in other cognates in which those consonants appear. #e may saely assume
that the non-1ermanic consonants are older than the 1ermanic ones. ?ence we
may accept with conidence &assuming a similar comparison o the vowels' the
reconstructions Moinos, Mdw, and Mtreyes as representing the Indo-European
orms rom which the existing orms developed. !omparative linguists have used
all the Indo-European languages as a basis or their conclusions regarding
correspondences,
not Gust the ew cited here.
I"2/E!$I5" I" $?E I"45-E6,5%EA" /A"16A1ES
All Indo-European languages are inlective@that is, all have a grammatical system
based on modiications in the orm o words, by means o inlections &endings and
vowel changes', to indicate such grammatical unctions as case, number, tense, person,
mood, aspect, and the like. Examples o such inlections in 0odern English are
cat[cats, mouse[mice, who[whom[whose, walk[walks[walked[walking, and sing[
sings[sang[sung[singing. $he original Indo-European inlectional system is very
imperectly represented in most modern languages. English, 2rench, and Spanish,
or instance, have lost much o the inlectional complexity that once characteri(ed
them. 1erman retains considerably more, with its various orms o noun, article,
and adGective declension. Sanskrit is notable or the remarkably clear picture it
gives us o the older Indo-European inlectional system. It retains much that has been lost
or changed in the other Indo-European languages, so that its orms show
us, even better than 1reek or /atin can, what the system o Indo-European must
have been.
Some 7erb Inlections
#hen allowance is made or regularly occurring sound changes, the relationship o
the personal endings o a verb in the various Indo-European languages becomes
clear. 2or example, the present indicative o the Sanskrit verb cognate with English
to bear is as ollows;
Sanskrit
bhar-mi <I bear=
bhara-si <thou bearest=
bhara-ti <he\she beareth=
bhar-mas <we bear=
bhara-tha <you &pl.' bear=
bhara-nti <they bear=
$he only irregularity here is the occurrence o -mi in the irst person singular,
as against -o in the 1reek and /atin orms cited immediately below. It was a peculiarity
o Sanskrit to extend -mi, the regular irst person ending o verbs that had no
vowel aixed to their roots, to those that did have such a vowel. $his vowel &or
example, the -a suixed to the root bhar- o the Sanskrit word cited' is called the
thematic vowel. $he root o a word plus such a suix is called the stem. $o these
stems are added endings. $he comparatively ew verbs lacking such a vowel in
Indo-European are called athematic. $he m in English am is a remnant o the
Indo-European ending o such athematic verbs.
/eaving out o consideration or the moment dierences in vowels and in initial
consonants, compare the personal endings o the present indicative orms as they
developed rom Indo-European into the cognate 1reek and /atin verbs;
1reek /atin
pherI erI
pherei-s er-s+
pherei- er-t
phero-mes &4oric' eri-mus
phere-te er-tis
phero-nti &4oric' eru-nt
!omparison o the personal endings o the verbs in these and other languages
leads to the conclusion that the Indo-European endings were as ollows &the Indo-
European reconstruction o the entire word is given in parentheses';
Indo-European
-, -mi &Mbher'
-si &Mbheresi'
-ti &Mbhereti'
-mes, -mos &Mbheromes'
-te &Mbherete'
-nti &Mbheronti'
1othic and early 5ld English show what these personal endings became in
1ermanic;
1othic Early 5ld English
bair-a ber-u, -o
bairi-s biri-s
bairi-Y biri-Y
baira-m bera-YI
bairi-Y bera-Y
baira-nd bera-Y
)2rom the earliest period o 5ld English, the orm o the third person plural was used throughout the plural.
$his orm, beraY, rom earlier MberanY, shows Anglo 2risian loss o n beore Y.
1ermanic Y &that is, TU' corresponds as a rule to %roto-Indo-European t.
/eaving out o consideration such details as the -nd &instead o expected -nY' in
the 1othic third person plural orm, or which there is a soundly based explanation,
the 1ermanic personal endings correspond to those o the non-1ermanic Indo-
European languages.
Some "oun Inlections
Indo-European nouns were inlected or eight cases; nominative, vocative, accusative,
genitive, dative, ablative, locative, and instrumental. $hese cases are modiications
in the orm o nouns, pronouns, and adGectives that show the relationship o
such words to other words in a sentence. $ypical uses o the eight Indo-European
cases &with 0odern English examples' were as ollows;
nominative; subGect o a sentence &$hey saw me.'
vocative; person addressed &5icer, I need help.'
accusative; direct obGect &$hey saw me.'
genitive; possessor or source &Shakespeare=s play.'
dative; indirect obGect, recipient &1ive her a hand.' ablative; what is separated &?e abstained rom
it.'
locative; place where &#e stayed home.'
instrumental; means, instrument &She ate with chopsticks.'
$he ull array o cases is preserved in Sanskrit but not generally in the other
descendant languages, which simpliied the noun declension in various ways. $he
paradigms in the ollowing table show the singular and plural o the word or
<horse= in %roto-Indo-European and ive other Indo-European languages. Indo-
European also had a dual number or designating two o anything, which is not
illustrated.
Indo-European "oun 4eclensionI
Indo-European Sanskrit 1reek /atin 5ld Irish 5ld English
Singular
"om. Mekwos avas hippos eCuus ech eoh
7oc. Mekwe ava hippe eCue eich
Acc. Mekwom avam hippon eCuum ech n-- eoh
1en. Mekwosyo avasya hippou eCu eich os
4at. Mekwy av]ya hippi eCu eoch o
Abl. Mekwd av]d eCu
/oc. Mekwoy ave
Ins. Mekw avena
%lural
".\7. Mekws av]s hippoi eCu eich os
Acc. Mekwons av]n&s' hippous eCus eochu os
1en. Mekwm av]n]m hippn eCurum ech n-- ona
4.\Ab. Mekwobh&y'os avebhyas hippois eCus echaib om
/oc. Mekwoysu avesu
Ins. Mekwys avais
I$here are a good many complexities in these orms, some o which are noted here. In 1reek, or the genitive singular,
the ?omeric orm hippoio is closer to Indo-European in its ending. $he 1reek, /atin, and 5ld Irish nominative
plurals show developments o the pronominal ending M-oi, rather than o the nominal ending M-s. !eltic was alone
among the Indo-European branches in having dierent orms or the nominative and vocative plural> the 5ld Irish
vocative plural was eochu &like the accusative plural', a development o the original nominative plural Mekws. $he
1reek and /atin dative-ablative plurals were originally instrumental orms that took over the unctions o the other
cases> similarly, the 5ld Irish dative plural was probably a variant instrumental orm. $he /atin genitive singular - is
not rom the corresponding Indo-European ending, but is a special ending ound in Italic and !eltic &5ld Irish eich
being rom the variant Mekw'.
-$he 5ld Irish n- in the accusative singular and genitive plural is the initial consonant o the ollowing word.
#5,4 5,4E, I" $?E I"45-E6,5%EA" /A"16A1ES
Early studies o the Indo-European languages ocused on cognate words and on
inlections. 0ore recently attention has been directed to other matters o the grammar,
especially word order in the parent language. :oseph 1reenberg &ESome
6niversals o 1rammarF' proposes that the orders in which various grammatical
elements occur in a sentence are not random, but are interrelated. 2or example,
languages like 0odern English that place obGects ater verbs tend to place modiiers ater
nouns, to put conGunctions beore the second o two words they connect, and
to use prepositions;
verb Y obGect; &$he workman' made a horn.
noun Y modiier; &$hey marveled at the' si(e o the building.
conGunction Y noun; &!ongress is divided into the Senate' and the ?ouse.
preposition Y obGect; &?arold ought' with him.
5n the other hand, languages like :apanese that place obGects beore verbs tend
to reverse the order o those other elements@placing modiiers beore nouns, putting
conGunctions ater the second o two words they connect, and using postpositions
&which are unction words like prepositions but come ater, instead o beore,
a noun'. 0ost languages can be identiied as basically either 75 languages &like
English' or 57 languages &like :apanese', although it is usual or a language to
have some characteristics o both types. English, or example, regularly puts adGectives
beore the nouns they modiy rather than ater them, as 75 order would
imply.
#inred %. /ehmann &%roto-Indo-European Syntax' has marshaled evidence
suggesting that %roto-Indo-European was an 57 language, even though the existing
Indo-European languages are generally 75 in type. Earlier stages o those languages
oten show 57 characteristics that have been lost rom the modern tongues
or that are less common than ormerly. 2or example, one o the oldest records o a
1ermanic language is a runic inscription identiying the workman who made a
horn about A.4. )**;
ek hlewagasti, holtiGa, horna tawido
I, ?lewagastir ?oltson, TthisU horn made.
$he order o words in sentences like this one &subGect, obGect, verb' suggests
that %roto-1ermanic had more 57 characteristics than the languages that evolved
rom it.
In standard 0odern 1erman a possessive modiier, as in der 1arten des
0annes <the garden o the man,= normally ollows the word it modiies> the other
order@des 0annes 1arten <the man=s garden=@is possible, but it is poetic and
old-ashioned. In older periods o the language, however, it was normal. Similarly,
in 0odern English a possessive modiier can come either beore a noun &an 57
characteristic', as in the building=s si(e, or ater it &a 75 characteristic', as in the
si(e o the building, but there has long been a tendency to avor the second order,
which has increased in reCuency throughout much o the history o English. In the
tenth century, practically all possessives came beore nouns, but by the ourteenth
century, the overwhelming percentage o them came ater nouns &.).) to IJ.^ percent,
,osenbach IL_'. $his change was perhaps under the inluence o 2rench,
which may have provided the model or the phrasal genitive with o &translating
2rench de'.
#hen we want to Goin two words in English, we put the conGunction beore the
second one &a 75 characteristic', as in the Senate and people. But /atin, preserving
an archaic eature o Indo-European, had the option o putting a conGunction ater
the second noun &an 57 characteristic', as in senatus populusCue, in which -Cue is a
conGunction meaning <and.= 0odern English uses prepositions almost exclusively,
but 5ld English oten put such words ater their obGects, so that they unctioned as
postpositions, thus;
?arold him wiB geeaht.
?arold him with ought.
Evidence o this kind, which can be ound in all the older orms o Indo-
European and which becomes more reCuent the arther back in history one
searches, suggests that Indo-European once ordered its verbs ater their obGects. I
that is so, by late Indo-European times a change had begun that was to result in a
shit o word-order type in many o the descendant languages rom 57 to 75.
0A:5, !?A"1ES 2,50 I"45-E6,5%EA" $5 1E,0A"I!
5ne group o Indo-European speakers, the 1ermanic peoples, settled in northern
Europe near 4enmark. 1ermanic dierentiated rom earlier Indo-European in the
ollowing ways;
I. 1ermanic has a large number o words that have no known cognates in
other Indo-European languages. $hese could have existed, o course, in
Indo-European but been lost rom all other languages o the amily. It is more
likely, however, that they were developed during the %roto-1ermanic period or
taken rom non-Indo-European languages originally spoken in the area occupied
by the 1ermanic peoples. A ew words that are apparently distinctively
1ermanic are, in their 0odern English orms, broad, drink, drive, owl, hold,
meat, rain, and wie. $he 1ermanic languages also share a common inluence
rom /atin, treated in !hapter I- &-).[_'.
-. 1ermanic languages have only two tenses; the present and the preterit &or
past'. $his simpliication o a much more complex Indo-European verbal
system is relected in English bind[bound, as well as in 1erman binden[band
and 5ld "orse binda[band. "o 1ermanic language has anything comparable
to such orms as those o the /atin uture, perect, pluperect, and uture
perect orms &or instance, laudb, laudv, laudveram, laudver', which
are expressed in the 1ermanic languages by verb phrases &or instance, English
I shall praise, I have praised, I had praised, I shall have praised'.
+. 1ermanic developed a preterit tense orm with a dental suix, that is, one
containing d or t &as in spell[spelled Tsp`ld, sp`ltU' alongside an older pattern o
changing the vowels inside a verb &as in rise[rose'. All 1ermanic languages
have these two types o verbs. 7erbs using a dental suix were called weak by
the early 1erman grammarian :acob 1rimm because they needed the help o a
suix to show past time. 7erbs that did not need such assistance, he called
strong. 1rimm=s metaphorical terminology is not very satisactory, but it is still
used. An overwhelming maGority o our verbs add the dental suix in the
preterit, so it has become the regular and only living way o inlecting verbs in
English and the other 1ermanic languages. All new verbs orm their preterit
that way; televise[televised, rev[revved, dis[dissed, and so orth. And many
older strong verbs have become weak. ?istorically speaking, however, the vowel change
in the strong verbs, called ablaut or gradation &as in drive[drove
and know[knew', was Cuite regular. 5n the other hand, some weak verbs,
which use the dental suix, are irregular. Bring[brought and buy[bought, or
instance, are weak verbs because o the suix -t, and their vowel changes do
not make them strong. "o attempt at explaining the origin o this dental suix
has been wholly satisactory. 0any have thought that it was originally an
independent word related to do.
). All the older orms o 1ermanic had two ways o declining their adGectives.
$he weak declension was used chiely when the adGective modiied a deinite
noun and was preceded by the kind o word that developed into the deinite
article. $he strong declension was used otherwise. $hus 5ld English had Y
geongan ceorlas <the young ellows &churls',= with the weak orm o geong, but
geonge ceorlas <young ellows,= with the strong orm. $he distinction is preserved
in present-day 1erman; die Gungen Aerle, but Gunge Aerle. $his particular
1ermanic eature cannot be illustrated in 0odern English, because English
has happily lost all such declension o adGectives. $he use o the terms strong
and weak or both verbs and adGectives, in Cuite dierent ways or the two
parts o speech, is unortunate but traditional.
J. $he EreeF accentual system o Indo-European, in which the accent shited
rom one syllable to another in various orms o a word, gave way to the
1ermanic type o accentuation in which the irst syllable was regularly stressed,
except in verbs like modern believe and orget with a preix, whose stress was
on the irst syllable o the root. "one o the 1ermanic languages has anything
comparable to the shiting accentuation o /atin vZr <men,= virarum <o the
men= or o hbbe <I have,= habcmus <we have.= !ompare the paradigms o the
1reek and 5ld English developments o Indo-European Mp tcr <ather=;
1reek 5ld English
Singular nominative patcr der
Singular genitive patrds der&es'
Singular dative patrZ der
Singular accusative patXra der
Singular vocative pbter der
%lural nominative patXres deras
%lural genitive patXrn dera
%lural dative patrbsi derum
%lural accusative patXras deras
In the 1reek orms, the accent may occur on the suix, the ending, or the
root, unlike the 5ld English orms, which have their accent ixed on the irst
syllable o the root. 1ermanic accent is also predominantly a matter o stress
&loudness' rather than pitch &tone'> Indo-European seems to have had both
types o accent at dierent stages o its development.
^. Some Indo-European vowels were modiied in 1ermanic. Indo-European o was
retained in /atin but became a in 1ermanic &compare /atin octo <eight,= 1othic ahtau'.
!onversely, Indo-European became 1ermanic &/atin mter
<mother,= 5E mdor'.
L. $he Indo-European stops bh, dh, gh> p, t, k> b, d, g were all changed in what is
called the 2irst Sound Shit or 1rimm=s /aw. $hese changes were gradual,
extending over long periods o time, but the sounds eventually appear in
1ermanic languages as, respectively, b, d, g> , , h> p, t, k.
2I,S$ S56"4 S?I2$
1rimm=s /aw
Because the 2irst Sound Shit, described by 1rimm=s /aw, is such an important dierence
between 1ermanic and other Indo-European languages, we illustrate it
below by &I' reconstructed Indo-European roots or words &or convenience omitting
the asterisk that marks reconstructed orms', &-' corresponding words rom a
non-1ermanic language &usually /atin', and &+' corresponding native English
words. &5nly a single Indo-European root is given or each set, although the ollowing
words may be derived rom slightly dierent orms o that root. $hereore, the
correspondence between the two derived words and the Indo-European root may
not be exact in all details other than the initial consonants.'
I. Indo-European bh, dh, gh &voiced stops with a pu o air or aspiration, represented
phonetically by a superscript TeU' became respectively the 1ermanic
voiced ricatives , B, f, and later, in initial position at least, b, d, g. Stated in
phonetic terms, aspirated voiced stops became voiced ricatives and then unaspirated
voiced stops. $hese Indo-European aspirated sounds also underwent
changes in most non-1ermanic languages. $heir developments in /atin, 1reek,
and 1ermanic are shown in the ollowing table;
Indo-European bh dh gh &that is, TbeU, TdeU, and TgeU'
/atin - - h- &initially> medially; -b-, -d- or -b-, -g-'
1reek &that is, TpeU, TteU, TkeU, transliterated ph, th, ch'
1ermanic b d g
Aeep these non-1ermanic changes in mind, or the ollowing examples will not
make sense;
Indo-European bh /atin -, 1reek ph 1ermanic b
bhrter rter brother
bhibhru- iber beaver
bhl lre blow
bhreg- ra&n'go break
bhudh- undus &or Mudnus' bottom
bhgo- gus beech
bhg- &1k.' phgein <to roast= bake
Indo-European dh /atin -, 1reek th 1ermanic d
dheigh- i&n'gere <to mold= dough
dhwer- oris door
dh- &1k.' th- <to place= do
dhug&h't9r &1k.' thugatr daughter
Indo-European gh /atin h-, 1reek ch 1ermanic g
ghordho- hortus &5E' geard <yard=
ghosti- hostis guest
ghomon- homo gome &obsolete, but
in brideg&r'oom'
ghol- &1k.' chol &g cholera' gall
ghed- &pre'he&n'dere <to take= get
ghaido- haedus <kid= goat
-. Except when preceded by s, the Indo-European voiceless stops p, t, k became
respectively the voiceless ricatives , , x &later h in initial position';
Indo-European p /atin, 1reek p 1ermanic
pt9r pater ather
pisk- piscis ish
pel- pellis ell <animal hide=
pr- &1k.' pr ire
prtu- portus ord
pulo- pullus oal
ped- ped&em' oot
peku- pecu <cattle= ee &c. 1er. 7ieh <cattle='
Indo-European t /atin t 1ermanic
treyes trs three
ters- torrre <to dry= thirst
t t &5E' Y <thou=
ten- tenuis thin
tum- tumre <to swell= thumb &that is, <at inger='
ton- tonre thunder
Indo-European k /atin k &spelled c, C' 1ermanic h
krn- corn horn
kerd- cord- heart
kwod Cuod what &5E hwht'
ker- cervus hart
kmtom cent- hund&red'
kel- cl]re <to hide= hall, hell
kap- capere <to take= heave, have
+. $he Indo-European voiced stops b, d, g became respectively the voiceless stops
p, t, k.
Indo-European b /atin, 1reek, /ithuanian,
,ussian b
1ermanic p
treb- trabs <beam, timber=
&g TarchiUtrave'
&archaic' thorp <village=
dheub- &/ith.' dubRs deep
abel- &,uss.' Gabloko apple
$he sound b was inreCuent in Indo-European and extremely so at the beginning o
words. Examples other than those above are hard to come by.
Indo-European d /atin, 1reek d 1ermanic t
dw duo two
dent- dentis tooth
dem- domre tame
drew- &1k.' drs <oak= tree
dekm decem ten &1othic taZhun'
ed- edere eat
Indo-European g /atin, 1reek g 1ermanic k
genu- genu knee &loss o Tk-U is
modern'
agro- ager <ield= acre
gen- genus kin
gwen- &1k.' gun <woman= Cueen
grno- grnum corn
gn- &g'nscere know, can
7erner=s /aw
Some words in the 1ermanic languages appear to have an irregular development o
Indo-European p, t, and k. Instead o the expected , , and x &or h', we ind , B,
and f &or their later developments'. 2or example, Indo-European ptr &represented
by /atin pater and 1reek patr' would have been expected to appear in 1ermanic
with a medial . Instead we ind 1othic adar &with d representing TBU', Icelandic
aB ir, and 5ld English hder &in which the d is a #est 1ermanic development o
earlier TBU'. It appears that Indo-European t has become B instead o .
$his seeming anomaly was explained by a 4anish scholar named Aarl 7erner
in I.LJ. 7erner noticed that the %roto-1ermanic voiceless ricatives &, , x, and s'
became voiced ricatives &, B, f, and (' unless they were prevented by any o three
conditions; &I' being the irst sound in a word, &-' being next to another voiceless
sound, or &+' having the Indo-European stress on the immediately preceding syllable.
$hus the t o Indo-European ptr became , as 1rimm=s /aw predicts it should> but then,
because the word is stressed on its second syllable and the is
neither initial nor next to a voiceless sound, that ricative voiced to B .
7erner=s /aw, which is a supplement to 1rimm=s /aw, is that %roto-1ermanic
voiceless ricatives became voiced when they were in a voiced environment and the
Indo-European stress was not on the immediately preceding syllable. $he law was
obscured by the act that, ater it had operated, the stress on 1ermanic words
shited to the irst syllable o the root, thus eectively disguising one o its important
conditions. &$he eect o the position o stress on voicing can be observed in
some 0odern English words o oreign origin, such as exert Tig(rtU and exist
Tig(istU, compared with exercise T`ksjrsai(U and exigent T`kskntU.' $he later history
o the voiced ricatives resulting rom 7erner=s /aw is the same as that o the
voiced ricatives that developed rom Indo-European bh, dh, and gh.
$he ( that developed rom earlier s appears as r in all recorded 1ermanic languages
except 1othic. $he shit o ( to r, known as rhotacism &that is, r-ing, rom
1reek rho, the name o the letter', is by no means peculiar to 1ermanic. /atin ls
<lower= has r in all orms other than the nominative singular@or instance, the genitive
singular lris, rom earlier Ml(is, the original s being voiced to ( because o
its position between vowels.
#e have some remnants o the changes described by 7erner=s /aw in present-day
English. $he past tense o the verb be has two orms; was and were. $he alternation
o s and r in those orms is a result o a dierence in the way they were stressed in
prehistoric times. $he 5ld English verb rosan <to ree(e= had a past participle rom
which came a now obsolete adGective rore <rosty, ro(en.= $he 5ld English verb
orlosan <to lose utterly= had a past participle rom which came our adGective orlorn.
Both these orms also show the s\r alternation. Similarly, the verb seethe had a past
participle rom which we get sodden, showing the T\dU alternation. In early
1ermanic, past participles had stress on their endings, whereas the present tense
orms o the verbs did not, and that dierence in stress permitted voicing o the last
consonant o the participle stems and hence triggered the operation o 7erner=s /aw.
$he SeCuence o the 2irst Sound Shit
$he consonant changes described by 1rimm and 7erner probably stretched over
centuries. Each set o shits was completed beore the next began and may have
occurred in the ollowing order;
I. Indo-European &IE' bh, dh, gh &respectively' 1ermanic &1mc' , B, f
-. IE p, t, k &respectively' 1mc , , x & h initially'
+. 1mc , , x, s &respectively' 1mc , B, f, ( &under the conditions o 7erner=s
/aw'
). IE b, d, g &respectively' 1mc p, t, k
J. 1mc , B, f, ( &respectively' 1mc b, d, g, r
#ES$ 1E,0A"I! /A"16A1ES
$he changes mentioned in the preceding section aected all o the 1ermanic languages,
but other changes also occurred that created three subgroups within the 1ermanic
branch@"orth, East, and #est 1ermanic. $he three subgroups are
distinguished rom one another by a large number o linguistic eatures, o which
we can mention six as typical;
I. $he nominative singular o some nouns ended in -a( in %roto-1ermanic@or
example, Mwula(. $his ending disappeared completely in #est 1ermanic &5ld
English wul' but changed to -r in "orth 1ermanic &5ld Icelandic ulr' and to
-s in East 1ermanic &1othic wols'.
-. $he endings or the second and third persons singular in the present tense o
verbs continued to be distinct in #est and East 1ermanic, but in "orth
1ermanic the second person ending came to be used or both;
5ld English 1othic 5ld Icelandic
bindest bindis bindr <you bind=
bindeY bindiY bindr <he\she binds=
+. "orth 1ermanic developed a deinite article that was suixed to nouns@or
example, 5ld Icelandic ulr <wol= and ulrinn <the wol.= "o such eature
appears in East or #est 1ermanic.
). In #est and "orth 1ermanic the ( that resulted rom 7erner=s /aw appears
as r, but in East 1ermanic it appears as s; 5ld English are <ear= and 5ld
Icelandic eyra, but 1othic auso.
J. #est and "orth 1ermanic had a kind o vowel alternation called mutation
&treated in the next chapter'> or example, in 5ld English and 5ld Icelandic,
the word or <man= in the accusative singular was mann, while the corresponding
plural was menn. "o such alternation exists in 1othic, or which the
parallel orms are singular mannan and plural mannans.
^. In #est 1ermanic, the B that resulted rom 7erner=s /aw appears as d, but it
remains a ricative in "orth and East 1ermanic; 5ld English hder, 5ld
Icelandic aBir, 1othic aB ar &though spelled adar'.
#est 1ermanic itsel was divided into smaller subgroups. 2or example, ?igh
1erman and /ow 1erman are distinguished by another change in the stop sounds@
the Second or ?igh 1erman Sound Shit@which occurred comparatively recently as
linguistic history goes. It was nearing its completion by the end o the eighth century
o our era. $his shit began in the southern, mountainous part o 1ermany and spread
northward, stopping short o the low-lying northernmost section o the country. $he
high in ?igh 1erman &?ochdeutsch' and the low in /ow 1erman &%lattdeutsch' reer
only to relative distances above sea level. ?igh 1erman became in time standard
1erman.
#e may illustrate the ?igh 1erman shit in part by contrasting English and
?igh 1erman orms, as ollows. In ?igh 1erman;
%roto-1ermanic p appears as p or, ater vowels, as &pepper[%eer'.
%roto-1ermanic t appears as ts &spelled (' or, ater vowels, as ss &tongue[8unge> water[
#asser'.
%roto-1ermanic k appears ater vowels as ch &break[brechen'.
%roto-1ermanic d appears as t &dance[tan(en'. $he !ontinental home o the English was north
o the area in which the ?igh
1erman shit occurred. But even i this had not been so, the English language
would have been unaected by changes that had not begun to occur at the time o
the Anglo-Saxon migrations to Britain, beginning in the ith century. !onseCuently
English has the earlier consonantal characteristics o 1ermanic, which it shares with
/ow 1erman, 4utch, 2lemish, and 2risian.
Because English and 2risian &the latter spoken in the northern 4utch province
o 2riesland and in some o the islands o the coast' share certain eatures not
ound elsewhere in the 1ermanic group, they are sometimes treated as an Anglo-
2risian subgroup o #est 1ermanic. $hey and 5ld Saxon share other eatures,
such as the loss o nasal consonants beore the ricatives , s, and Y, with lengthening
o the preceding vowel; compare ?igh 1erman gans with 5ld English gs
<goose,= 5ld ?igh 1erman im &0odern 1erman ln' with 5ld English <ive,=
and ?igh 1erman mund with 5ld English mB <mouth.=
English, then, began its separate existence as a orm o 1ermanic brought by
pagan warrior-adventurers rom the !ontinent to the relatively obscure island that
the ,omans called Britannia and, until shortly beore, had ruled as part o their
mighty empire. $here, in the next ive centuries or so, it developed into an independent
language Cuite distinct rom any 1ermanic language spoken on the !ontinent.

You might also like