l2l - lectures on germanic
TRANSCRIPT
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 1/17
L2L History and structure of Western European languages
The Germanic languages – a synchronic and diachronic overview
1. External history
NB: Keep in mind that much of the early history of the Germanic languages is still
uncertain (for a detailed discussion see Nielsen’s book on the reading list).
It is thought that around 1000 BC the Southern Baltic region (Danish isles, southern
Sweden) was settled by Indo-European speakers. They may have encountered non-IE
speakers there and, perhaps as a result of language contact, changed their Proto-Indo-
European language into the earliest form of common Germanic, called Proto-
Germanic. There are no recorded data for Proto-Germanic, it is entirely a
reconstructed language.
The idea that there was contact between immigrating IE-speakers and original non-IE
speakers in the region and period just mentioned, and that Germanic arose as a result
of this contact, is seriously uncertain. The archaeological evidence on which this
hypothesis is based is open to various interpretations (Nielsen pp.35-36). It is based
on different ways of burial. The older way of burial (supposedly of the non-IE
speakers) was initially supposed to have co-existed with the newer type (supposedly
of the IE speakers). However, later evidence suggests that the new type is really from
a later date. This means that, instead this being an indication of IE speakers invading a
terrain of others, it may well be that we are dealing with the same group of people that
just started to bury their dead in a different way at some point in time.
Whatever the origins of Proto-Germanic, the speakers of this language at some point
began to migrate. Migration leads to contact with different languages, and to
separation of the migrating group from other speakers of Proto-Germanic. As a
consequence of both these factors, Proto-Germanic developed in different ways in
different areas, resulting in different branches of the Germanic language family. It is
usually assumed that three branches emerged, as indicated in the tree in (1). (See
Nielsen for a detailed discussion on how these branches relate to each other and on
whether this idea makes sense to begin with).
(1) Proto-Germanic
------------------------------------
------------------ East Germanic
West Germanic North Germanic
East Germanic has Gothic (now extinct) as its main representative. The North-
Germanic branch comprises the Scandinavian languages (Icelandic, Norwegian,
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 2/17
Faroese, Swedish, Danish), the remaining ones (German, Dutch, English, Frisian,
Yiddish, perhaps Afrikaans) belong to the West-Germanic branch.
Note that the idea that the history of languages can be represented by tree diagrams
such as (1) is problematic (cf. Nielsen, chapter V). It presupposes that there is a point
at which one language is split into two distinct ones, that these languages are isolatedfrom one another after the split and do not influence each other after this point, and it
ignores the influence of contact with other languages (i.e. the possibility that
languages may not have just one mother (proto-)language, but can have various
parents). This is unlikely in many cases. In fact, a very important trigger for language
change is contact with different languages.
(A result of contact with other languages is also that languages that are spoken within
a particular geographical area tend to start sharing certain features, a phenomenon
known as Sprachbund . In such a case, these shared features do not indicate that these
languages are genetically related. A famous example of this phenomenon today is the
Balkan Sprachbund, which comprises (a.o.) Bulgarian, Macedonian (both Slavic),Romanian (Romance), Albanian and Greek. Compare also the idea that there is a
Celtic substrate influence in particular branches of Germanic and Romance (French) –
if correct, this shows that non-related languages can come to share some features
because they have both superseded the same language in their areas).
Despite these caveats, (1) reflects at least part of what happened in the history of
Germanic. Let us see how migration lead to the development of these three different
branches (insofar as it did – we must also take care not to simply translate
archaeological evidence for certain tribal migratory movements into evidence for a
split of the Germanic languages along exactly the same lines).
The first tribe to move from the Germanic ‘homeland’ (i.e. the Danish isles, Southern
Sweden) were the Goths, around 100 BC. (They were later followed by other tribes
from different regions, such as the Burgundians and the Vandals). They first moved
eastward (hence the name of the branch, East Germanic), at least as far as the Black
Sea. A variant of Gothic, or at least some East Germanic language, was spoken in the
Crimea until around the 18th century. Later the Goths split up into two large groups,
the Visigoths (or West- Goths) and the Ostrogoths (or East-Goths). The Visigoths
migrated westward (to Italy, France, Spain), where they assimilated with Romance
speakers. All forms of Gothic are now extinct.
Gothic is important for linguists because the oldest texts we have of a Germanic
language are in Gothic. These are bible translations attributed to bishop Wulfila, from
around the mid of the 4th century AD, fragments of which are known to us from later
(6th c.) manuscripts. To get an idea of what the language looks like, here is the
paternoster (the Lord’s Prayer, ‘Our Father’) in Gothic:
(2) Atta unsar þu in himinam,
father ours you in heavens
weihnai namo þein.
be-hallowed name your
qimai þiudinassus þeins.may-come kingdom yours
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 3/17
Wairþai wilja þeins,
may-become will yours
swe in himina jah ana airþai.
as in heaven also on earth
Hlaif unsarana þana sinteinan gif uns himma daga.bread ours the daily give us this day
Jah aflet uns þatei skulans sijaima,
and forgive us that guilty may-be
swaswe jah weis afletam þaim skulam unsaraim.
like also we forgive the debtors ours
Jah ni briggais uns in fraistubnjai,
and not bring us in testing
Ak lausei uns af þamma ubilin.
but deliver us from the wicked-one
Amen
Some striking characteristics of Gothic are the following. (NB: Keep in mind that
Gothic is not an ancestor of the West and North Germanic languages – see the tree in
(1). However, given the proximity in time of the Gothic texts to Proto-Germanic it is
reasonable to assume that in the oldest stages of the other Germanic languages, for
which we have scarcely any data, they shared many of these characteristics).
• Inflectional endings still contain a full vowel (cf. below on deflexion), witness thecase ending in unsar , the plural/case ending in himinam, the optative ending in
weihnai, etc.
•There are four cases; there are several different declination classes for the nouns(apparently based on the form of the stem in Proto-Germanic), as well as three
genders.
• The definite article still has the same form as a demonstrative pronoun; there is no
indefinite article.
• The paradigms for personal pronouns distinguish not only singular and plural forms,
but also dualis forms, that is, forms referring to speaker and hearer together.
• There is a synthetic passive, that is, a passive that is made by only conjugating the
verb, rather than also adding an auxiliary, e.g. daupjada ‘(he) is being baptized’.
(However, in the preterite tense the passive already is analytic, e.g. daupiþs ist ‘he
was baptized’).
• There are some occurrences of ‘pro drop’, the phenomenon that pronominal subjectsneed not be realized syntactically (compare e.g. modern Italian), as in …þatei skulans
sijaima ‘… what (we) may be guilty of’, where the subject ‘we’ is not realized.
(However, it is difficult to draw conclusions about Gothic syntax from the texts, as
Wulfila seems to have followed the Greek original rather strictly in this respect.)
• There are traces of morphological reduplication as a means to form the preterite
tense of some verbs, e.g. haitan – haihait ‘to call’ – ‘called’.
The North Germanic branch remained uniform for a relatively long period (well into
the Middle Ages). Thus, the oldest attested forms of this branch, attested in runic
inscriptions (cf. week 1), are usually all referred to as Common Scandinavian, or
Common Norse. Toward the end of the Viking era, however, three distinct power
centres had emerged: Norway, Denmark and Sweden, and Common Scandinavian had
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 4/17
split into two different branches: East Scandinavian (or East Norse), spoken in
Denmark, the southern part of Sweden and the south-eastern part of Norway, and
West Scandinavian (or West Norse), spoken in most of Norway and the Norse
settlements in the Atlantic, in particular Iceland.
During the Middle Ages, East Scandinavian splits into Old Danish and Old Swedish(and Old Gutnish, a language spoken on Gotland). West Scandinavian comprises Old
Norwegian and Old Icelandic (the differences between these two are very minor).
Icelandic developed in relative isolation (and in a socially very cohesive society) after
the settlement, and is conservative compared to the other Scandinavian languages (it
has retained, for example, a much richer inflectional system, cf. below). Modern
Faroese also developed out of the West Scandinavian branch.
By way of illustration, (3) contains the paternoster in Old West Norse, and (4)
contains an Old Danish sentence.
(3) Faþer vár es ert í himenríki, verði nafn þitt hæilagtTil kome ríke þitt, værði vili þin
sva a iarðu sem í himnum.
Gef oss í dag brauð vort dagligt
Ok fyr gefþu oss synþer órar,
sem vér fyr gefom þeim er viþ oss hafa misgert
Leiðd oss eigi í freistni, heldr leys þv oss frá öllu illu.
(4) Kurmr konukr karþi kubl þusi aft þurui kunu sina tanmarkaR but.
Gorm king made monument this afterThorvi wife his Denmarks adornment.
(Stone from Jelling, 10th c.)
West Germanic is also supposed to be the result of certain tribes moving away from
the common Germanic ‘homeland’, amongst others the Frisians, Angles, Saxons,
Franks, Alemannians and Bavarians. Very roughly speaking, Old Frisian developed
out of the language of the Frisians, Old English out of the language of the Angles and
the Saxons, Old Dutch out of the language of the Franks, Old Low German out of the
language of the Saxons and the Franks, Old High German out of the language of the
Alemannians, the Bavarians and the Franks.
It is in fact rather unclear for the West Germanic languages that they all derive from
the same homogeneous parent language (i.e. the evidence for grouping them togetheris weaker than it is for North Germanic and East Germanic). East Germanic branched
off the earliest from the other languages because of the emigration of the Goths, but it
is not so clear when a common West Germanic ancestor branched off from North
Germanic. If the early runes (from the 3rd c. onward) are anything to go by, West and
North Germanic were still pretty uniform in the early Middle Ages – the runes show
elements that are common to all Germanic dialects except Old High German (the
southernmost Germanic language) and Gothic.
It is sometimes suggested that early West Germanic really consisted of three different
dialect groups (the names of which are taken from the Roman historian Tacitus, who
wrote a description of Germania): (i) Ingwaeonic (or North Sea Germanic) (ii)
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 5/17
Istwaeonic (or Rhine-Weser Germanic) (iii) Herminonic (or Elbe Germanic). It is
questionable, too, whether this is a useful distinction (cf. Nielsen).
Language contact played an important role, again, in the different later developments
of the various West Germanic languages (the development of English, for example,
was influenced by the Viking and Norman invasions).
Below are bits of Old Dutch (or Old Low Franconian) (5), Old High German (6), Old
Low German (or Old Saxon) (7), Old English (8), and Old Frisian (9).
(5) Forchta in biuonga quamon ouer mi in bethecoda mi thuisternussi
fear and trembling came over me and covered me darkness
In ic quad uuie sal geuan mi fetheron also duuon in ic fliugon sal in raston sal
and I said who shall give me feathers like dove and I fly shall and rest shall
(from the so-called Wachtendonck Psalms, original from 10th century)
(6) Ik gihorta ðat seggen, ðat sih urhettun ænon muotin, I heard that say that themselves warriors alone engaged
Hiltibrant enti Hadubrant untar heriun tuem
Hiltibrant and Hadubrant between armies two
sunufatarungo iro saro rihtun.
son-and-fathers their armour arranged
garutun se iro guðhamun, gurtun sih iro suert ana,
readied they their battle-coverings girded themselves their swords on
(from the Hildebrandslied , ca. 800)
(7) Fadar ûsa firiho barno,
thu bist an them hôhon himila rîkea,geuuîhid sî thîn namo uuordo gehuuilico.
Cuma thîn craftag rîki.
Uuerða thîn uuilleo obar thesa uuerold alla,
sô sama an erðo, sô thar uppa ist
an them hôhon himilo rîkea.
(beginning of the paternoster, from Heliand , 9th c.)
(8) Fæder ure þu þe eart on heofenum
sig þin nama gehalgod
to becume þin rice
geweorðe þin wylla
on eorðan, swa swa on heofenum
urne dæghwamlican hlaf syle us todæg
forgif us ure giltas
swa swa we forgifað urum giltendum
ne gelæde þu us on costnunge
ac alys us of yfele. Soðlice
(paternoster, 11th c.)
(9) Thit was to there stunde
it was at the time tha thi kening kerl riuchta bi gunde.
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 6/17
That the king Charles rule started
Tha waster ande there saxinna merik
Then was-there in the Saxon’s district
liudingerus en hera fele steric.
Liudingerus a lord much strong
(beginning of epic poem from the Hunsinger Codex, 13th c.)
2. Distinguishing features
What is the evidence that the Germanic languages form one family? In other words,
what makes a Germanic language a Germanic language and sets it apart from other
Indo-European languages?
Probably the most famous distinguishing feature is that all Germanic languages have
undergone a particular sound change, as compared to Proto-Indo-European. A sound
change here is to be understood as a diachronic change in the phonology of a
language. Sound changes can be schematized as follows:
(10) X→ Y / -- Z -- (at t)
which is to be read as ‘sound X changes into sound Y in the phonological context Z
(at time t)’. According to the so-called Neogrammarians of the end of the 19th c. (cf.
week 2), sound changes are exceptionless and do not work haphazardly: in all words
where sound X occurs in the context Z it changes into Y. This fits well into the ‘tree
model’ of language families: we can distinguish a clear split between a parent
language which has the original sound and a distinct daughter where the sound change
has worked in all words where it could have worked. (In fact, later research has
shown that sound changes typically gradually work their way through the lexicon of a
language, a process known as ‘lexical diffusion’. Crucially, it has also been found
that, when a sound change occurs, we can distinguish a central dialect in which it
starts and in which most words will be affected by it, and more peripheral dialects in
which it can affect some but not all relevant words, contrary to what the
Neogrammarians assumed. Compare the discussion of the ‘wave theory’ in Nielsen,
chapter V).
The sound changes that set Germanic apart from other Indo-European languages are
the ones in (11), collectively known as ‘Grimm’s law’ (after Jakob Grimm, a 19th c.
German philologist of folktale fame - although the relevant correspondences were
originally noticed by the Danish linguist Rasmus Rask). It is also referred to as the(first) Germanic Sound Shift.
(11) p, t, k → f, , X (the sound of ch in e.g. German rauchen ) (which in turn
could turn into h)
b, d, g → p, t, k
bh, dh, gh → b, d, g
(11) states that voiceless plosives became fricatives, voiced plosives became
voiceless, and aspirated plosives became de-aspirated. Grimm’s law is supposed to
have worked at around the 5th c. BC (it did not work in Latin loan words from a later
date – compare the [k] in cannabis with the [h] in hemp). Its effects are illustrated in(12).
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 7/17
(12) Latin piscis, versus English fish, German Fisch, Dutch vis
Sanskrit pita, Lat. pater, versus English father , German Vater , Dutch vader
Latin tenuis, versus English.thin
Latin centum (with initial [k] sound), but English Hundred , German hundert
Latin decem, versus English ten, Dutch tien, German zehn (cf. below)Sanskrit dva, versus English two, Dutch twee
Latin genus, versus English knee, Dutch knie
Sanskrit bhratar , versus English brother, German Bruder
There are apparent exceptions to Grimm’s law. These were explained by the
Neogrammarian Karl Verner. According to Verner’s law, voiceless fricatives,
including those that resulted from Grimm’s law, subsequently became voiced
obstruents when they followed an unstressed syllable. This law changed f, , X, and s
after an unstressed syllable into b, d, g and z, respectively. This law operated on the
basis of the old stress pattern; the stress shift that took place in Germanic (cf. below)
must have taken place after this sound shift.
A second striking innovation in the Germanic languages is the development of a
dental suffix to form the past tense. Compare the regular past tense formation in the
modern Germanic languages:
(13) Eng. work work-ed
Ger. arbeiten ‘to work’ arbeite-te ‘worked’
Du. werken ‘to work’ werk-te ‘worked’
Swe. kasta ‘to throw’ kasta-de ‘threw’
Fri. meane ‘to mow’ mean-de ‘mowed’
There is ‘much controversy’ (Nielsen p.30) about the origin of this suffix, but
possibly it originally was a form of the verb ‘to do’ (‘he work did’, i.e. ‘he worked’),
which got encliticized and then became a suffix. In Gothic we still find strong
remnants of the older system of past tense marking, which was by ablaut . This term
designates a vowel change in the stem to express a particular morphological category,
in this case past tense. There are five different patterns of such vowel change in
Gothic (so-called ablaut series); which one a particular verb belongs to depends on the
original form of its root. A remnant of this older system of past tense marking in the
modern Germanic languages are the so-called ‘strong’ verbs, such as English to
throw, with past tense threw. Many verbs have become ‘weak’, which means theyhave been made to take the new dental suffix as past tense marker in the course of
time.
NB: Do not confuse ablaut with umlaut . Umlaut is a vowel change in the stem that
does not express a particular morphological category, but which is phonologically
conditioned. It is a form of vowel harmony: the vowel in the stem adapts itself to
match the place of articulation of a vowel in a suffix that is added to the stem. This
process, too, played an important role in the history of the Germanic languages. A
form of umlaut is one feature which sets North and West Germanic apart as a group
from East Germanic (Gothic). It involves fronting of back vowels if the following
syllable contained an i (a front vowel). Often such cases of umlaut are no longertransparent because the suffix with the triggering vowel subsequently was lost. The
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 8/17
result may be that there is seemingly arbitrary allomorphy, that is, two forms for the
same stem (one which originally preceded the suffix with the conditioning vowel and
one which did not). Thus, the result of the mentioned form of umlaut that took place
in all Germanic languages except Gothic are pairs such as English foot - feet and
mouse - mice, where the plural stem forms originally had a plural suffix with a front
vowel following. Because this suffix was subsequently lost, this now looks much likea form of ablaut, as the only indication of plurality is the vowel change in the stem.
The next common Germanic property relates to word stress. Word stress in Proto-
Indo-European was variable (it depended on the inflectional class that a word
belonged to). In Proto-Germanic, the stress shifted to the initial syllable of the word. It
is sometimes suggested that this was the result of contact with speakers of a non-Indo-
European language (compare section 1 above), but this is difficult to prove. The result
of this stress shift was that the inflectional endings of words came to stand in a
phonologically weak position. This had consequences for the further diachronic
development of the Germanic languages, as discussed below.
A final Germanic innovation to be mentioned here is the development of a so-called
‘weak’ declension for attributive adjectives, which came into being next to the
‘strong’ declension that goes back to the Proto-Indo-European declensional system.
The weak declension is used in definite contexts. We can see the difference in modern
German pairs such as ein gut-er Mann (strong) versus der gut-e Mann (weak).
3. Internal history – some general developments
In this part we will discuss some general characteristics of the development of the
Germanic languages in the stages of which we have (non-reconstructed) data: how did
‘Old’ English/Frisian/Danish/Dutch etc. develop into ‘Middle’ English/Danish etc.
and how did these develop into the ‘(Early) Modern’ languages. Note that, in the
usual divisions into periods, the terms ‘Old’ and ‘Middle’ do not denote exactly the
same period for each of the respective languages. The table in (14) lists the divisions
in periods for the various languages. The exact dividing line between any two periods
is of course to some extent arbitrary, so you will find other divisions than the ones in
(14) in the literature as well.
(14) Old Middle (Early) Modern
English 8th-11th c. 11th-15th c. >15th c.
Frisian 11th-16th c. 16th-18th c. >18th c.
Dutch 9th-12th c. 12th-15th c. >15th c.Low German 8th-11th c. 11th-15th c. >15th c.
High German 8th-11th c. 11th-15th c. >15th c.
Danish/Swedish 7th-12th c 12th-15th c. > 15th c..
Icelandic 7th-16th c. -- . > 16th c.
Norwegian 7th-14th c. 14th.-16th c. > 16th c.
The distinction between ‘Old’ and ‘Middle’ can be based on, for example, whether or
not the inflectional endings of words still had a full vowel or were already reduced
(cf. below), but it can also simply reflect a gap in the textual evidence. For example,
there is a large time gap between the few Old Dutch fragments that exist and the
earliest Middle Dutch texts from the end of the 12th century. The beginning of the
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 9/17
Early Modern periods can be based on the beginning of some form of standardization
of the language.
To get an idea, the oldest texts we have of the various languages are from the
following periods (compare these dates with those of the Gothic texts, which go back
to a 4th c. original): Old English 8th c. (maybe late 7th c.); Old Frisian 11th c. (inmanuscripts from the 13th c.); Old Dutch 9th c.; Old Low German 9th c.; Old High
German 8th c.; Old East Scandinavian 13th c. (but many earlier runic inscriptions),
Old West Scandinavian 12th c. (but many earlier runic inscriptions).
NB It is important to realize that, nowadays, when we talk about ‘English’, ‘German’
or ‘Swedish’, and make distinctions between these languages, we usually refer to the
so-called standard variant of these languages and ignore the dialectal differences that
exist. However, in the Middle Ages the process of standardization had not taken place
yet. This means that when we talk of Old and Middle English, Old and Middle Dutch,
etc., we refer to a group of dialects, not to one and the same language that was spoken
in the whole area covered by the term ‘Middle Such-and Such’ (let alone in the areawhere Such-and-Such is currently the standard language). These dialects differ in a
number of respects (especially phonologically, but also morphologically and
sometimes syntactically), but they share enough structural traits to be classified under
the same header. It is important to realize as well that there still is considerable
dialectal variation today, next to the standard versions of the languages. Thus, as you
can see from the table, in the older stages of the languages a distinction is made
between Middle High German, Middle Low German and Middle Dutch. Nowadays
we distinguish German and Dutch. Does this mean that a whole language has
disappeared? No - many dialects that are descendants from Middle Low German are
still spoken today (‘Plattdeutsch’), but as it happened the area where they are spoken
lacked the political/economical power centre to impose its language as a standard
variant in the area. Modern standard German is more or less based on High German
dialects, so usually when people talk about ‘German’ they refer to the version of those
dialects that happened to have become the accepted standard variant for the whole of
Germany.
Here, then, are some striking characteristics of the development of the Germanic
languages from the early Middle Ages until now.
For a start, Proto-Indo-European is thought to have been a heavily inflecting language
(cf. week 2). The Germanic languages have undergone (are undergoing) a process of deflexion: they lose inflectional morphology. Deflexion can be observed in the history
of many Indo-European languages, but, with some exceptions, the Germanic
languages have been strongly influenced by this process. This may be related to the
stress shift mentioned above. When stress was shifted to the initial syllable of words,
the inflectional endings came to stand in a phonologically weak position in the word.
The result was that the full vowels in the endings could reduce to schwa, leading to
paradigm levelling (the use of the same suffix for different functions within a
morphological paradigm). Subsequently, inflectional suffixes could get lost entirely.
On the nominal side of things, deflexion manifests itself for instance in a loss of
morphological case markers. Thus, modern English, Dutch and MainlandScandinavian have lost morphological case altogether, whereas the older stages of
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 10/17
these languages still had case (though in the recorded stages for the most part it
already did not show up on nouns themselves anymore; instead, it showed up on
determiners and attributive adjectives):
(15) case endings for sg. masc. NPs in ME MDutch Old Norse
Nominative ∅ ∅ -rGenitive -es -s -s
Dative -e -n -i
Accusative ∅ -n ∅
Modern German still has some morphological case, though the paradigm has become
poorer in its distinctions here, too. Icelandic still has a very rich case system, basically
identical to that of Old Norse.
Another instance of deflexion in the nominal system is the loss of morphological
gender. Whereas modern German still distinguishes three genders, modern Dutch only
distinguishes two, and modern English one.
On the verbal side of things, we can see a loss of agreement morphology (for person
and number). Compare the following Middle Dutch and Old Norse paradigms for
person/number agreement on verbs with their present-day counterparts:
(16) Old Norse Modern Norwegian Middle Dutch Modern Dutch
1sg ∅ -r ∅ ∅
2sg -r -r -s -t
3sg -r -r -t -t
1pl -um -r -en -en2pl -ið -r -t -en
3pl -a -r -en -en
Most of the languages also show loss of verbal inflection for mood, in particular, loss
of a distinct subjunctive ending.
The degree to which the various languages have been subject to deflexion differs.
Icelandic still has rich inflectional paradigms, German is somewhat ‘in between’,
Mainland Scandinavian, English and Dutch are strongly deflected. It has been
suggested that there is a connection between the degree of contact with other
languages and the degree of deflexion: compare the relative isolation of Iceland withthe seafaring/trading/colonizing history of England, Holland and Scandinavia. Most
strongly deflected is Afrikaans, which has almost no inflectional morphology at all,
and which was developed in a situation of ‘severe’ language contact.
Directly related to deflexion is the fact that the languages become less synthetic and
more analytic in nature. Roughly speaking, we can divide languages into ‘cut’
languages (analytic) and ‘paste’ languages (synthetic). In the latter, grammatical
categories such as the perfect or the passive or the future tense are formed by
morphologically altering the word to which they apply, for example by adding an
affix to it. In the former, such categories are made by using separate auxiliary words,
such as auxiliary verbs or modals. Compare the Latin perfect in (17) with the present-day Italian one (the Romance languages, too, have become more analytic).
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 11/17
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 12/17
(22) a. .. þæt he mot ehtan godra manna (VO)
that he might persecute good men
b. a se Wisdom þa þis fitte asungen hæfde … (OV)
when the Wisdom then this poem sung had
In fact, the order that predominated in Old English was OV, so English is often said to
have changed from an OV to a VO language. Crucially, however, we find instances of
both orders in the documented older stages of the language. When we look at the
history of what is still an OV language today, Dutch, we see that, despite the
prevalence of OV orders in Middle Dutch, here too we find VO orders as well in the
older stages:
(23) a. T’ Allen denghenen die dese letteren selen sien (OV)
to all those that these letters will see
b. …dat hi sal senden sine ghesele den sondaren (VO)
that he will send his whip the sinners
Similar word order freedom can be detected for some other constituents as well.
Note that this does not mean that in the older stages of the languages ‘anything goes’.
The languages were certainly different from what they are today, but not more
‘simple’ in any meaningful sense (compare the greater word order freedom, which
might be considered ‘simpler’, with the presence of the complex inflectional system
that has now been lost).
An important phonological change, which sets off Old High German on the one hand
from Old Low German/Old Dutch on the other hand, is the so-called High German
Sound Shift, also known as the Second Germanic Sound Shift (a somewhat unlucky
term, since in contrast to the first sound shift, i.e.Grimm’s Law, it did not affect all the
Germanic languages). This involved the following changes:
(24) p → pf (→ f)
t → ts (→ s)
k → kX (→ X)
In words, voiceless plosives become affricates (plosive plus fricative), or
(subsequently) just fricatives. This accounts for present-day contrasts such as the
following:
(25) standard German Dutch English
Pfeffer peper pepper
Waffen wapen weapon
Zeit [tsait] tijd time
Wasser water water
Buch [buX] boek book
This change is believed to have taken place somewhere between the 5th and the 8th
centuries. It started in the south, in the mountainous area of Germany – hence the
name of the dialects that were most strongly affected by it: High German. From thereit spread northwards, its effect becoming ever weaker. In the low parts of the country
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 13/17
the dialects were not affected by it, hence the term Low German for these dialects.
Dialects in the middle could be affected to some extent: they can simultaneously have
words that show the effect of the sound shift and words that do not. Thus, we can
draw so-called isoglosses that indicate the dividing line between the two different
pronunciations for a particular word. The dividing line between High and Low
German is sometimes taken to be the maken – machen isogloss, also known as the‘Benrather Linie’ (after the town through which it goes).
4. Some further structural properties of the Modern Germanic languages
We have already encountered a few structural properties of the Modern Germanic
languages. In this final part we will discuss two more striking grammatical features
that are found in some or many of them.
The first has to do with the positioning of the finite verb in a clause. Considering
English, we see that in a declarative clause the finite verb must follow the subject
(26a). It does not matter whether we put another constituent than the subject in the
first position, the finite verb still follows the subject (compare (26b) with (26b’)).
(26) a. Mary will go to the theatre tonight.
b. Tonight Mary will go to the theatre.
b’. *Tonight will Mary go to the theatre
In questions, however, this is different. Here, the finite verb has to shift to a position
in front of the subject (unless the subject itself is questioned – question words have to
be in the very first position of the clause). There is so-called subject-verb inversion:
(27) a. Have you seen that man before?
b. Will she go to Denmark?
c. What has Mary read today?
d. Where can we go?
In English, only auxiliaries/modals invert with the subject in questions, main verbs
cannot do this. If a sentence with just a main verb is questioned, the dummy auxiliary
do must appear as finite verb, in the inverted position:
(28) a. Karin read the paper.
b. *What read Karin?
c. What did Karin read?
In the other Germanic languages, however, main verbs can undergo the same
movement to the inverted position that auxiliaries undergo in English questions.
Moreover, all these languages have a much more pervasive use of the ‘verb shifting’
rule than English, where it is restricted to questions. As a first approximation, we may
say that the finite verb moves to the left in all main clauses in these languages. To be a
bit more precise, the verb moves to the position that comes second in the sentence’s
structure, since there can be no more than one constituent in front of the finite verb.
(In some constructions the finite verb comes first, such as in ‘yes/no’ questions, cf.
English (27a-b). In these cases, the first position in the sentence structure is
apparently left empty). The phenomenon therefore is known as verb second , or V2.All Germanic languages except English are verb second languages.
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 14/17
Consider for example the difference with respect to the position of the finite verb
between the main clauses and the embedded clauses in the following examples from
Dutch (exactly parallel examples can be given for German, Frisian and Afrikaans).
(29) a. (Ik geloof) dat Marie de krant leest. I think that Mary the paper reads
a’. *(Ik geloof) dat Marie leest de krant.
b. Marie leest de krant.
Mary reads the paper
b’. *Marie de krant leest.
Whereas the embedded clause shows subject-object-verb order, the main clause has
subject-verb-object order. At first sight, this might seem to imply that Dutch is a
‘typologically mixed’ language where it concerns the order between the verb and its
complement, in other words, a ‘mixed’ OV/VO language. Perhaps V takes its
complement to the left in embedded clauses, but to the right in main clauses. A closerlook at all relevant data shows, however, that this is not an attractive hypothesis. For a
start, in main clauses with a periphrastic predicate, it is only the finite verb that shows
up in front of the object; all other verbs obligatorily follow it (i.e. they appear in OV
order), see (30). Contrast this with the word order in a true VO language such as
English, as illustrated by the translations in (30).
(30) a. Marie heeft het boek gelezen.
Mary has the book read
‘Mary has read the book.’
a’. *Marie heeft gelezen het boek
b. Karel zou die film hebben willen zien.
Carl would that film have want see
‘Carl would have wanted to see that film’
b’. *Karel zou hebben willen zien die film.
Moreover, it can be shown in a straightforward way that at least the finite verb is not
in its ‘normal’ position in any sentence that has a constituent other than the subject in
the first position. This is because in such sentences the finite verb inverts with the
subject, just as in the English questions in (27), hence it must be in a position higher
than the subject position here.
(31) a. Gisteren las Marie de krant.
yesterday read Mary the paper
‘Yesterday Mary read the paper’.
a’. *Gisteren Marie las de krant.
b. Volgens mij heeft ze die film niet gezien.
according me has she that film not seen
‘I think she has not seen that film’.
b’. *Volgens mij ze heeft die film niet gezien.
On the basis of data as in (30) and (31) people have concluded, then, that a language
like Dutch is really an OV language, which has a general rule that shifts the finiteverb to second position in main clauses. In addition to this, one constituent, no matter
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 15/17
which one, can be put into to the first position in main clauses, thereby giving these
clauses their ‘verb second’ character.
It is possible to be somewhat more precise about what the ‘second position’ in a
clause is. People have argued that it is the position in which we see complementizers
showing up in embedded clauses. Evidence for this assumption comes from theobservation that the distinction between clauses with V2 and clauses without V2 is
actually not exactly a main clause – embedded clause distinction, as assumed so far. A
more precise generalization is that V2 occurs in any clause without a complementizer .
Conversely, in any clause that contains a complementizer, V2 is blocked. There are
non-embedded clauses that contain a complementizer, such as the exclamation in
(32). As you see, the finite verb cannot be fronted in such a clause. Conversely, there
are embedded clauses without a complementizer, such as the ‘reported speech’ clause
in (33), and these show fronting of the finite verb. Particularly revealing are certain
conditional embedded clauses, as in (34). These can be introduced by the
complementizer als, in which case the finite verb stays at the end of the conditional
clause, as in (34a). However, these particular embedded clauses need not beintroduced by a complementizer, but in that case the finite verb moves to the front, as
in (34b). (In this case, as in ‘yes/no’-questions, there is no constituent in the first
position in front of the moved verb).
(32) a. Een lekkere dingen dat ik gisteren in dat restaurant gegeten heb!
A tasty things that I yesterday in that restaurant eaten have
‘I had incredibly good food in that restaurant yesterday!’
b. *Een lekkere dingen dat ik heb gisteren in dat restaurant gegeten!
(33) a. Hij zei [hij had de krant al gelezen]
He said he had the paper already read
‘He said he had already read the paper’
b. *Hij zei [hij de krant al gelezen had]
(34) a. [Als hij nog komen mocht] dan geef je hem maar een biertje
if he yet come might then give you him but a beer-DIM
‘In case he yet comes, just serve him a beer.’
b. [Mocht hij nog komen], dan geef je hem maar een biertje
‘idem’
This shows that if there is no complementizer, the finite verb moves into the‘complementizer position’. If this position is filled already, the finite verb stays in its
basic position. To make things more complicated, there are in fact two Germanic
languages that do show general V2 in embedded clauses that contain an overt
complementizer. Embedded V2 of this type occurs in Icelandic (34a) and Yiddish
(34b).
(34) a. (Jón harmar) að þessa bók skuli ég hafa lesið
Jon regrets that this book should I have read
b. (…) az morgn vet dos yingl zen a kats
that tomorrow will the boy see a cat
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 16/17
Interestingly, languages that have basic VO-order, like English, can also show Verb
Second. The Scandinavian languages exemplify this. That these languages have the
V2 rule is shown by the fact that, as in Dutch and unlike in English, if any constituent
other than the subject appears in the first position in a (main) clause, the finite verb
must invert with the subject. The following Danish examples show this (see also the
embedded V2 case in the VO language Icelandic just mentioned in (34a)). Theembedded clause in (35a) illustrates the VO-character of Danish. (35b)-(35d) show it
has V2. Contrast (35c)-(35d) with English this film the children have not seen versus
*this film have the children not seen.
(35) a. … at børnene har set filmen
that children- DEF have seen film- DEF
‘…that the children have seen the film’
a’. * … at børnene filmen har set
b. Børnene har set filmen.
c. Denne film har børnene set.
this film have children- DEF seen ‘The children have seen THIS film’.
d. * Denne film børnene har set.
So having the Verb Second rule or not is independent from being a VO language or an
OV language. (Note that there are many OV languages that do not show V2 outside
Germanic, for example Turkish or Japanese).
A second property of some Germanic languages that is, as far as we now know, rather
unique, is that they show complementizer agreement . In the relevant varieties, not
only finite verbs show person and number agreement with the subject of the clause,
but so do complementizers. The phenomenon occurs, amongst others, in Flemish,
Bavarian, Frisian, and in some East-Netherlandic dialects. The data in (36) illustrate
the phenomenon for West Flemish.
(36) a. dan ik werken
that-1sg I work-1sg
b. da gie werkt
that -2sg you work -2sg
c. da ze/Vale`re werkt
that -3sg she / Valerie work -3sg
d. dan wunder werkenthat -1pl we work -1pl
e. da gunder werkt
that -2pl you.pl work -2pl
f. dan zunder/Pol en Vale`re werken
that-3pl they/Paul and Valerie work-3pl
In West Flemish, the morphological paradigm for complementizer agreement is
arguably identical to that for finite verbs, as in (37).
(37) -en (1sg), -t (2sg), -t (3sg), -en (1pl), -t (2pl), -en (3pl)
7/27/2019 L2L - Lectures on Germanic
http://slidepdf.com/reader/full/l2l-lectures-on-germanic 17/17
The –t ending is truncated before a consonant (as in (36b, c, e) and voiced before a
vowel (so if the subject begins with a vowel we see dad ),
Some East-Netherlandic dialects are even more spectacular, in that they have a
paradigm that is especially used for complementizer agreement. Thus, in the
embedded clause in (38) (from the Hellendoorn dialect in the region of Overijssel) thesubject agrees with both the complementizer and the finite verb, but the agreement
ending on the former (-e) is different from that on the latter (-t ).
(38) a. …datte wij speult
that-pl we play-pl
a’. …*datte wij speule
Interestingly, in main clauses where the finite verb inverts with the subject, the finite
verb must show the special ending from the complementizer-agreement paradigm; in
main clauses where the subject comes first this is not so, see (39) vs. (40). Apparently
the morphology is sensitive to whether or not the subject follows or precedes theagreeing element.
(39) a. Wij speult.
we play-pl
b. *Wij speule.
(40) a. Dan speule wij.
then play-pl we
b. *Dan speult wij.