Professional Documents
Culture Documents
The most decisive change in the establishment of the correlation of syllable cut in
the Germanic languages, and hence in the loss of phonological quantity, was the
vowel lengthening that took place in the middle periods of the Germanic languages.
It did away with the VC type of Old Germanic. This is, at the same time, the most
widespread of the quantity changes contributing to the correlation of complemen-
tary length, which is still in evidence in parts of Germanic territory. The quantity
changes in the Old Germanic languages treated in Chapter 2 were not as wide-
spread geographically or did not have as great an effect on the vocabulary of the
languages in question as did the vowel lengthening to be discussed presently.
Rivaling the importance of vowel lengthening in Middle Germanic is perhaps
the final strengthening discussed in Chapter 2. Much of the evidence for this change
in the old periods of the Germanic languages was to be done away with by vowel
lengthening in monosyllables. This latter change is usually taken to be merely
a footnote to vowel lengthening in open syllables (OSL), which has attracted, and
still attracts, the lion’s share of attention in linguistic research. Yet in certain
respects, monosyllable lengthening (MSL) is still rather poorly understood. One
open question is the relative chronology of OSL and MSL.
The tasks of the present chapter will be to deal with the most important factors
involved in the lengthening process throughout Germanic, including the role of
consonants, to establish the main language-specific issues involved in the
lengthening process, and to survey the geographic distribution of open syllable
lengthening and monosyllable lengthening in the Modern Germanic languages
and dialects. At first, we will examine the common nature of the process in
Germanic as whole, before we turn our focus to the individual areas.
70
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.1 The Cause and Environment of Lengthening 71
the change. Although it may seem rather simple in retrospect to separate out
factors of environment from the cause of the change, this has not always been
done in research on the topic. Oftentimes, scholars were content to establish the
phonetic law by which vowels were lengthened. This task necessarily involves
stating the environment in which the change occurred, but all too often such
a law has been seen as explaining or “motivating” the change. As will be seen,
an explanation of the change, which also explains why the change took place
when it did, is to be preferred over explanations that are divorced from the rest
of language history and can be plugged in equally well (or equally poorly) into
another period.
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
72 Arriving at the Goal: Vowel Lengthening in Middle Germanic
The logic behind viewing the open syllable as the environment for OSL
is that the vowel is allowed to lengthen in the absence of a following
tautosyllabic consonant preventing its extension. Compare the use of the
terms “free” versus “checked” vowels in English. After all, final vowels in
monosyllabic words were “free” to lengthen in the old periods of the
Germanic languages (cf. Section 2.3.1). The rule of the open syllable
could of course not apply to MSL. A sound law requiring lengthening in
open syllables fails, because there is great variation before sonorants in the
following syllables (Liberman 1992: 68–69). Compare examples like Eng.
liver, oven, hammer; Ger. Leber, Ofen, Hammer; Du. zeven (“seven”), oven,
hamer. From the overview of the variation present in the Modern Germanic
languages given below, it will be clear that there was merely a strong
tendency to vowel lengthening, rather than a law governing its presence.
It is obvious from the writings of Paul and Luick that the position of the
vowel in the open syllable was to them merely the environment in which
the change occurred. As was discussed in Section 1.3, they both sought the
cause of the change in Sievers’ syllable accents. Especially Luick, but also
Paul, thought the language was striving to reach a normal measure (Ger.
normales Maß) for all words. This concept was discussed in detail in
Section 1.1.3.
Most language histories and historical grammars of Middle English, Middle
High German, etc. written since the days of the Neogrammarians thus refer to
the change as “open syllable lengthening”. For English, see Morsbach (1896:
84), Luick (1914–40, 1: 379–99), Wyld (1921: 112–13), Wright and Wright
(1928: 40–44), Jordan (1934: 44–46), Brunner (1938: 17 and 1960: 105–08),
Horn and Lehnert (1954, 1: 163–66), Prins (1972: 105–08); or for German, see
Moser (1929: 73–76), Priebsch and Collinson (1934: 153–56), Kienle (1969:
37–40), Bach (1970: 227–28), Paul and Klein (2007: 82). For Dutch, Frisian,
and Scandinavian, see Franck (1910: 12–14), Schönfeld and Loey (1959:
32–33), Loey (1964: 30–31 and 1967: 36), Goossens (1974: 42–43), Van
Loon (1986: 86–90); Steller (1928: 115–16), Bremmer (2009: 115–16);
Noreen (1913: 117, 141; 1970: 110), Torp and Falk (1898: 17–24), Jacobsen
(1910: 42–43), Brøndum-Nielsen (1950: 379–81), Hansen (1962–71: 333–36).
Some of these works do not look at the change analytically, which may be
appropriate depending on the scope of the work, but it is fair to say that it has
been taken for granted that the environment for OSL is the open syllable, hence
the conventional name for the change. It is sometimes argued, however, that the
open syllable is an important factor, if not the most important factor in the
change (cf. Weinstock 1975; Hubmayer 1982; King 1988). Relatively recent
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.1 The Cause and Environment of Lengthening 73
works by Lahiri and Dresher (1999) and Becker (2002) explicitly attribute the
lengthening to the open syllable.
Yet it is questionable whether the open syllable is the environment for the
lengthening, since we cannot be certain about Old or early Middle Germanic
syllable structure. It is entirely possible that the syllables in question were not
open, but closed. This is the view that will be argued here. How could speakers
of Middle English or its contemporaries pronounce syllables which could not
be pronounced as words? Anatoly Liberman (1992: 77) is to be credited with
bringing this point to bear on the discussion of OSL. For him, the syllable is an
autonomous phonetic unit, equal to a monosyllabic word, and division into
syllables is equivalent to division into phonetic words. In this line of thinking,
he follows Jerzy Kuryłowicz (1948a), who in turn harkens back to an older
tradition. In German scholarship at the turn of the century, for example, Gustav
Burghauser (1891) questioned whether these syllables were open or closed.
Liberman notes that Modern English has la-ter and lay, let-ter and let, while
early Middle English has nose (“nose”), speken (“speak”), but not *nŏ, *spĕ.
Such “over-short” words were already non-existent in the old periods by the
posited final vowel lengthening (cf. Section 2.3.1). So it is much more likely
that ME disyllables such as these had closed accented syllables, which became
open after the vowels were lengthened. The term open syllable lengthening
would thus be something of a misnomer. It is retained here out of convention
and for convenience sake.
One question related to the interplay between OSL and MSL is the relative
chronology of the two changes. If MSL is owing to analogy to OSL, as in the
majority view, then there is no question about which change is original. Yet
there is a theory that posits MSL before OSL. This view has its roots in the older
Scandinavian research on the topic (see, for example, Kock 1882–86, 1882:
382–92; Hesselman 1901; cf. Hansen 1962–71: 388–89). More recently, the
priority of MSL over OSL has been advocated for by Larissa Naiditsch and
Jurij Kusmenko (1992; see also Kusmenko 1995; Naiditsch 1997; and Page
2001). They seek the key to understanding vowel lengthening in the relic
dialects of Scandinavian (Swedish, Norwegian) in the extreme north and
Switzerland in the extreme south, which have MSL, but lack complete OSL
(for examples, see Section 3.2.1). These areas still retain segmental quantity
and, apparently, mora counting. The main motivation for vowel lengthening in
this view is the tendency to isochrony in the Germanic languages (see
Section 1.1).
Kusmenko and Naiditsch argue that the internal chronology of vowel length-
ening, and potential consonant lengthening, depends on the phonetic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
74 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.1 The Cause and Environment of Lengthening 75
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
76 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.1 The Cause and Environment of Lengthening 77
“ice”) first lost the secondary accent and later the unaccented -e, then com-
pensated for its loss through circumflex of the accented vowel, which in turn
allowed modulation and diphthongization of the accented vowel. Change in
the non-apocopated forms and monosyllables would have followed by
analogy.
In more recent times, the idea of compensation in OSL was reawakened in
a passing comment by Alice Grundt (1976: 10–13), who mixes and matches
a variety of approaches in her work. A fuller treatment is applied by Donka
Minkova (1982 and 1985). She attributes OSL to the principle of preservation
of the overall rhythmical weight of the foot (cf. Section 1.1.3). Her statistical
study of Holthausen’s (1934) etymological dictionary of Old English and Bliss’
(1955) list of Anglo-Norman words borrowed before 1400, all containing the
correct environment (open syllable), showed that 84 percent of the potential
bases for OSL have OSL in Modern English. She concludes that OSL operates
unfailingly only when there is schwa deletion, the loss of the syllable being
compensated for by the acquisition of a new mora in the accented syllable, thus
preserving the total number of morae and the overall duration of the word.
Roger Lass (1985) essentially agrees with Minkova, but compares the change
to metrical resolution, in that a disyllabic foot with two light syllables becomes
a monosyllabic foot with one heavy syllable.
Yet as Liberman (1992: 68–74) points out, it is not so much a mechanical
compensation for the loss of an unaccented vowel that triggers OSL, but for the
loss of the vowel’s function: the compensation is primarily functional and only
secondarily physical. He reasons that words like OE nosu had two vocalic
morae, each of which could serve as the locus of sentence stress. But through
apocope, endings began to lose their importance. As the full vowels were
replaced by schwa or lost, the root vowel began to perform the function
previously divided between the two morae: it became bimoric. The change in
the accented vowel, Liberman reasons, could be manifested physically by
roughly doubling in length, diphthongizing, or becoming circumflected.
In one way or another, phonological information is transferred to the accented
syllable, a trend which runs like a red thread throughout Germanic language
history. Words ending in resonants could protect themselves against apocope
by syllabifying the sonorant when the vowel in the unaccented syllable was
lost. As a compensatory change, OSL does not automatically result from the
weakening of full vowels, but its occurrence is present in all Germanic lan-
guages, even if it is not consistent.
The results of the last Germanic apocope vary from language to language.
Yet this poses no problem for the compensation theory. Following Bengt
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
78 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 79
name evidence in Dutch (Loey 1964: 30–31; Van Loon 1986: 86–90), doubling
of vowel graphs or accent marks in Danish law texts (Diderichsen 1937–38:
156–57; Haugen 1976: 258–59; Hansen 1962–71, 1: 334–35), at least in
monosyllables. The attestations vary quite a bit throughout Germanic territory,
but the orthographic evidence may appear considerably later than the actual
change in spoken language, perhaps after 100 years or even later. If the dating
for the different languages based on this type of information is correct, it would
mean generally that the change began in the central part of Germanic territory
and spread in all directions, with Upper German and Icelandic being the last
languages reached. The Scandinavian areas with restoration of full vowels,
Swedish and Icelandic, would also be among the last areas to acquire OSL.
Such a model of spread would also generally be reinforced by the location of
the relic areas on the edges of Germanic territory (see Section 3.2 for more
details). Hogg (1996) and Liberman (1997: 120–21) argue that there may have
been traces of MSL already in Old Scandinavian, Old English, and Old High
German, but not on nearly the same scale as at the time of OSL.
Based on the connection to apocope and the direct spelling evidence, open
syllable lengthening can be linked to the very transition from the old periods to
the middle periods of the Germanic languages. Thus, along with the loss of full
vowels in unaccented syllables as a primary phonological distinction between
Old High German and Middle High German, on the one hand, and Old English
and Middle English, on the other, we must also include the onset of open
syllable lengthening.
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
80 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 81
In Swedish, the areas where VC remains include the East Swedish dialects
in Finland and Estonia, a part of Norrland, and areas in Dalarna and western
Uppland. Vowel length in East Swedish has been studied very well by
Hultman (1894), Hesselman (1902), and Bergroth (1922) and is discussed
in descriptions of the following dialects: Gamlakarleby (Hagfors 1891:
16–20) and Vörå (Freudental 1889: 40–41) in Österbotten, Pargas
(Thurman 1898: 12–13) and Nagu (Thurman 1929: 72–75) in Åboland,
Nyland (Wessman 1945: 15–16) on the southern coast of Finland, and Runö
(Vendell 1882: 27–28) and Nuckö (Danell 1905–34: 208) in Estonia.
The quantity conditions of these dialects are most often described as being
similar to Old Swedish, i.e., without OSL or MSL. Åland (Hesselman 1902:
14–15) and Kokär in Åboland (Karsten 1892: 83–95), on the other hand, have
OSL and MSL. The vowel lengthening which is most prevalent is MSL,
especially of OSw. a. Several of the dialect descriptions mention shortening
in old VːCː sequences, by consonant shortening in Gamlakarleby, or by vowel
shortening in Nagu and Nyland. In Nuckö, there was some consonant length-
ening following short vowels.
In Sweden proper, lack of vowel lengthening is reported for several areas in
Norrland (Söderström 1972): Kalix and Piteå in Norrbotten, Nordmaling in
Västerbotten, and Ragunda in Jämtland. For the first three areas on the Gulf of
Bothnia, it is reported that short vowels remain in disyllables with Accent 2, for
Ragunda that OSw. i, u (o) remain short in open syllables. Yet OSw. a, æ are
often lengthened, especially before OSw. g/ɣ. There is more MSL than OSL in
these areas, yet short i, u (o) remain, especially before l, r. In Piteå, there is
lengthening of m following short vowels. For Överkalix (Kettunen 1983:
46–50) in Norrbotten, it is reported that short syllables and overlong syllables
remain, but are rather rare.
Finally, there are Central Swedish areas with short syllables in Dalarna and
western Uppland and bordering eastern Västmanland. Dialect descriptions are
available from Upper Dalarna (Levander 1925–28, 1: 60, 75ff) in northwestern
Dalarna, Orsa (Boëthius (1918:26: 13–14) in northeastern Dalarna,
Dalabergslagen (Envall 1930–47: 66ff, 71ff) in southeastern Dalarna, and
Fjärdhundraland (Isaacson 1923: 35–40) in southwestern Uppland. In these
dialects, short vowels remain in open syllables, but are usually lengthened in
monosyllables. They would be the best evidence for the primacy of MSL over
OSL. Compare examples from Upper Dalarna: ɡopɔ (Sw. gapa; “gape”), ipin
(öppen; “open”), pepor (peppar; “paper”), ɡata (gata; “street”), tjitil (kittel;
“kettle”), boko (baka; “bake”), wriden (vriden; “twisted”), fluɡu (fluga; “fly”),
luɡi (låga; “flame”), with variation from town to town. Yet there is some
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
82 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 83
data upon which the chronology of the changes in Germanic as a whole can be
based. It is difficult to determine the origin of the forms in these relic areas in
retrospect. Perhaps MSL simply spread into these areas from areas with OSL,
while OSL did not spread there.
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
84 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
Figure 3.2 Final Strengthening/no Monosyllable Lengthening in Bavarian (Base Map After Kranzmayer 1956, Map 22).
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
86 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Lusern (Bacher 1905) in Italy and the former speech island of Zarz (Lessiak
1959: 83) in present-day Slovenia, high valleys in Tirol (Kranzmayer 1960),
and the Lesach Valley (Lessiak 1903) in Carinthia. In these areas, final
strengthening, in the form of lengthening, remains (see Section 2.3.2 for
more discussion), thus preserving the state from the Old Germanic period,
which first established complementary length in monosyllables. In many of the
same areas, over-long syllables remain: the (former) speech islands Zarz
(Lessiak 1959), Lusern (Bacher 1903), and Sieben and Dreizehn Gemeinden
(Kranzmayer 1981), high valleys in Tirol (Kranzmayer 1960: 178–79), and
South Tirol (Schatz 1903: 70) generally. Compare examples found in the high
valleys of Tirol: hɔːɐssn̩ (Ger. heißen; “to call, name”), ʒloːuffn̩ (schlafen;
“sleep”), ʃproːxxe (Sprache; “language”) raːxxn̩ (rauchen; “smoke”),
kxroːuppfn̩ (Krapfen; “doughnut”), hɔːɐttsn̩ (heizen; “to heat”).
The number of relic dialects with respect to vowel lengthening and the
establishment of complementary length is relatively small in Bavarian in
comparison to Alemannic, which is generally more conservative than
Bavarian, especially Middle Alemannic and South Alemannic in southern
Baden-Württemberg, Switzerland, Italy, and Vorarlberg in Austria. There is
a large area in the central and southern part of Middle and South Alemannic,
which lacks complete OSL. Lengthening before obstruents is the exception,
while there is common but irregular lengthening of e, a, o before sonorants,
especially r. Many of these dialects have been described: Schwenningen (Haag
1898: 15; Bohnenberger 1900: 142–43), southern Saulgau (1932: 5–10),
Jestetten (Keller 1963: 73) near Waldshut, Schaffhausen (Stickelberger 1881:
410–16, Wanner 1941: 68–72), Singen (Schreiber 1928: 14–15), Zürcher
Oberland (Weber 1923: 74ff), Sense- and southeastern Seebezirk (Henzen
1927: 100ff) and Jaun (Stucki 1917: 127–46) in Freiburg, Goldbach
(Haldimann 1903: 337–38) in Upper Aargau, Entlebuch (Schmid 1915:
110ff) and the city of Luzern (Fischer 1960: 26–27) in Luzern, Brienz
(Schild 1891: 23) in Bern, Kerenzen (Winteler 1876: 82–84) in Glarus
(Streiff 1915: 48–52), as well as those listed below. Compare examples from
Entlebuch with lengthening before sonorants, but seldom before obstruents,
and then only of a, æ: ʃpaːrə (Ger. sparen; “save, spare”), faːrə (fahren; “go,
travel”), wɛːrə (wehren; “restrain”), b̥ ɛːri (Beere; “berry”), maːlə (malen;
“paint”), waːlə (sich wälzen; “wallow”), maːnə (mahnen; “remind”), faːnə
(Fahne; “flag”), raːmə (Rahmen; “frame”), b̥ ræːmə (Bremse; “horse-fly”), but
b̥ irə (Birne; “pear”), ʃtælə (stehlen; “steal”), ʃalə (Schale; “peel, bowl”), himəl
(Himmel; “sky, heaven”), namə (Name; “name”), hanə (Fasshahn; “tap”), ʃinə
(Schiene; “track”); versus saːɡə (sägen; “to saw”), træːɡ̊ ə (tragen; “carry”),
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 87
naːse (Nase; “nose”), but æb̥ ə (eben; “even”), sib̥ ə (sieben; “seven”), ʃad̥ ə
(Schaden; “damage”), fad̥ ə (Faden; “thread”), læd̥ ər (Leder; “leather”), lɛɡ̊ ə
(legen; “lay”), ʃwiɡ̊ ərə (Schwiegermutter; “mother-in-law”), xæfə (Erbse;
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
Figure 3.4 Final Strengthening/no Monosyllable Lengthening in Alemannic (Base Map After Bohenenberger 1953).
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 89
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
90 Arriving at the Goal: Vowel Lengthening in Middle Germanic
exception. Leo Jutz (1931: 156–57) described the areas as the southwest and
the northeast, the latter area including Kesswil and South Swabian east of the
line Engen-Tuttlingen-Friedingen. From these areas, there are several dialect
descriptions: Brienz in Bern (Schild 1891: 23), Canton Schaffhausen (Wanner
1941: 68–72), Wangen and the Höri area (Singer 1965: 61–63) near Konstanz,
Liggersdorf (Dreher 1919: 12–13) north of Konstanz, Kesswil (Enderlin 1913:
65–66) in Upper Thurgau, and the Rhine Valley (Berger 1913: 82–84) in
St. Gallen. The following examples are from Kesswil: taːl (Ger. Tal; “valley”),
tsaːl (Zahl; “number”), waːl (Wahl; “choice”), mɛːl (Mehl; “flour”), hoːl (hohl;
“hollow”), βøːl (wohl; “well”), stiːl (Stiel; “handle”), tsiːl (Ziel; “goal”), b̥ aːr
(bar; “bare”), ɡ̊ æːr (gar; “done”), fɔːr (vor; “in front of”), tɔːr (Tor; “gate”),
moːr (mürbe; “mellow, well cooked”), tsɔː (Zahn; “tooth”), weː acc. (wen
“who”), d̥ eː acc. (den “who”), eː acc. (ihn “he”), troːm (MHG trum, Ger.
Fadenende; “thread end”), but lamː (lahm; “lame”), d̥ emː dat. (dem “who”),
emː dat. (ihm “who”) versus ɡrab (Grab; “grave”), ʃtab̥ (Stab; “staff”), rad̥
(Rad; “wheel”), ɡ̊ led̥ (Glied; “limb”), wɛɡ̊ (Weg; “way, path”), tsuɡ̊ (Zug;
“train”), ɡ̊ ræs (Gras; “grass”), mɔs (Moor; “bog, moor”).
The South and Middle Alemannic dialects confront us with the same ques-
tion as the Scandinavian relic dialects with extensive MSL, but no OSL. Do the
conditions in these dialects not speak for the origin of vowel lengthening in
monosyllables? Here we deal with a larger area, but the problem is not as great,
since OSL is present, albeit to a limited degree, and could be a source for
analogy in the monosyllables with later generalization. As we have seen,
a number of dialects have restricted MSL with the same scope as OSL in
most of the dialects of the area, namely before sonorants, but not obstruents.
These dialects could speak for the original analogy of MSL to OSL in all South
and Middle Alemannic dialects lacking complete OSL. Analogy as the source
for the High Alemannic lengthening has previously been proposed by Virgil
Moser (1929: 73–76) in his Early New High German grammar and Viktor
Schirmunski (1962: 184) in his handbook of German dialects.
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 91
those without and those with MSL. Both of these types are quite pre-
valent. It is argued that the former group without MSL shows the earlier
stage of development and retains the previous stage of development for
the latter group with MSL by analogy to OSL, however long it may have
lasted.
Although there are no North Germanic dialects of this type, there is a large
area in continental West Germanic and a few smaller areas with complete OSL,
but no corresponding MSL. The result is paradigmatic variation between
disyllabic forms with long vowels and monosyllabic forms with short vowels.
The large area consists of Dutch, Low German, and Frisian dialects. Standard
Dutch belongs to this type. There are also several Upper German dialects that
belong here. Generally stated, these dialects have OSL in disyllables and retain
the state from Old Germanic in monosyllables with final strengthening.
Together, these two changes account for the (former) correlation of comple-
mentary length in these dialects. The state of these numerous dialects presents
important evidence for the priority of OSL over MSL and for the origin of MSL
via analogy to OSL.
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
92 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 93
“late”), mɔːɡŋ̩ (MLG maken; “make”), khnɔːɡŋ̩ (MLG knoke, Ger. Knochen;
“bone”), ɔːbm̩ (MLG open; “open”), khɔːm̩ (OS kuman, MLG komen; “come”),
fɔːʒl̩ (OS fugal, MLG fogel, Ger. Vogel; “fowl”), ɛːzl̩ (MLG esel, Ger. Esel;
“donkey, ass”), lɛːm̩ (MLG leven, Ger. leben; “live”), lɛːn (OS hlinon, MLG
lenen; “lean”), vɛːdn̩ (OS witan, MLG weten, Ger. wissen; “know”), œːvl̩ (OS
ūbil, MLG ovel, Ger. übel; “evil”). No lengthening has occurred in monosyl-
lables: bat (MLG bat; “bath”), fak (MLG vak, Ger. Fach im Schrank; “com-
partment”), blek (MLG blek, Ger. Blech; “sheet of metal”), vex (MLG wech;
“way”), bit (MLG bit, Ger. Gebiss am Zaum; “bit”), dik (MLG dick, Ger. Dir,
dich; “thee”), ɡot (MLG got, Ger. Got; “God”), rok (OS rok, Ger. Rock;
“skirt”), not (MLG not; “nut”).
Based on the short diphthongs in Westfalian, Agathe Lasch (1914, 1915,
1974: 35–38; see also Priebsch and Collinson 1934: 153–56) made the proposal
that the Middle Low German Tondehnung of short vowels in open syllables was
in fact by way of diphthongization, or as she called it, Zerdehnung. She
hypothesized that the short vowels first became short diphthongs (e > éè),
which are still present today in Westfalian, and that the modern dialects show
a variety of reflexes, following dissimilation and accent shift. While most
dialects would have lengthened and monophthongized the diphthongs, the
long diphthongs in open syllables, still attested in southern Brandenburg,
would have been reflexes of the intermediate stage after lengthening.
The meager evidence of diphthongization presented from Middle Low
German manuscripts by Lasch is hardly convincing and she has received little
support for her view (cf. Frings 1915; Foerste 1957: 1771–72). Nor has
Sarauw’s idea that the Westfalian short diphthongs are derived from long
diphthongs. It hardly seems that diphthongization is the general mechanism
for OSL, even in Low German. The Brandenburg diphthongs must be the result
of secondary diphthongization after lengthening. Generally speaking,
diphthongization is connected with long vowels and short diphthongs are
a rare phenomenon, yet probably also the result of Old English breaking.
Long vowels have the duration that allows modulation during the course of
their articulation, while short vowels do not.
Open syllable lengthening in Frisian is not much discussed. The change is
dated rather late, by continental West Germanic standards. It is assigned to the
fourteenth century in late Old Frisian (Hofmann 1969; Bremmer 2009:
115–16). This means it is attested in the Old West Frisian texts, which are
later than the Old East Frisian texts. Dieter Hofmann points to the doubling of
consonants following short vowels as an orthographic sign that segmental
quantity has been lost through OSL. By all accounts, OSL was less regular in
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
94 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Old Frisian than in English, Dutch, and Low German. The handbooks (Siebs
1901: 1179ff; Steller 1928: 8–12; Bremmer 2009: 115–16) differ regarding the
scope of the lengthening. Rolf Bremmer, for example, writes that OFris. i,
u were always lengthened, a, o frequently, and e not at all. Several (Van Helten
1890: 43–45; Heuser 1903: 6–8; Sjölin 1969: 19–21, 29–30) do not posit OSL
for Old Frisian, but this may be due to their focus on Old East Frisian.
Lengthening seems to be more regular in Modern Frisian dialects. Examples
from the North Frisian dialect of Wiedingharde (Jensen 1925: 46ff) on the west
coast of Schleswig-Holstein will give some impression of the scope: thraːvə
(LG draven, Ger. traben; “trot”), draːbə (cf. Go. ga-draban, Ger. treffen; “hit,
meet”), væːda (OFris. weder; “weather”), sæːʒl̩ (OE segl; “sail”), phlæːʒə
(OFris. plegia, Ger. pflegen; “take care of”), hɔːbə (OFris. hopia; “hope”),
nɔːz (OFris. nose; “nose”), khnɔːkə (MLG knoke, Ger. Knochen; “bone”), leːvə
(OFris. libba, Ger. leben; “live”), leːvərə (OFris. livria, Ger. liefern; “deliver”),
føːʒl̩ (OFris. fugel, Ger. Vogel; “fowl”), tøːʒl̩ (LG tögel, Ger. Zügel; “rein,
bridle”). In monosyllables, the vowels remain short: dax (OFris. thach, Ger.
doch; “yet”), vas (OFris. was; “was”), hem (OFris. him; “him”), khen (OFris.
kin, Ger. Kinn; “chin”), hof (OFris. hof, Ger. Hof; “courtyard”), khokh (MLG
kok; “cook”), jokh (OS juk; “yoke”), plom (LG plum; “plum”).
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 95
(Mehl; “flour”), b̥ ær (Bär; “bear”), lam (lahm; “lame”). So despite the High
German Consonant Shift, these southern Upper German dialects have nearly the
same quantity type as Dutch, Low German, and Frisian dialects in the northern part
of West Germanic. The differences that go along with OSL without MSL in these
two areas will be discussed in the upcoming chapters.
3.2.3.1 English
Vowel lengthening in Middle English has been studied extensively, first during
the 1890s. After a lengthy break, interest has been steady since 1950, with
a second flowering 100 years after the first, during the 1990s. The present
survey of limited scope will thus not turn up anything new about the geography
or the systematic nature of OSL in English. The change is dealt with in some
detail in several handbooks of Middle English and histories of English (Sweet
1888: 167–68; Morsbach 1896: 84–92; Luick 1914–40: 397–408; Wright and
Wright 1928: 40–44; Jordan 1934: 44–46; Berndt 1960: 25–28). The basic facts
are relatively clear after over a century of research on the topic.
The low and mid vowels, OE a, e, o, were lengthened to ME ā, ę̄ , ǭ over the
entire English-speaking territory, while OE i, u were lengthened to ME ē, ō
primarily only in the north, including the Scots, Northern English, and northern
Midlands areas. The lowering of the high vowels in connection with lengthening
is shared, as we saw, with Dutch and Low German. The controversy of the
geographic extent of the lengthening of the high vowels was put to rest through
the work of Karl Luick (1896, 1899a, 1899b, 1903; cf. Morsbach 1897; Sarrazin
1890 and 1898). Outside of the northern area just mentioned, based on the data of
Ellis’ dialect survey, he noted traces of lengthening in a large area in the south
encompassing northeastern Somerset, Wiltshire, Oxfordshire, Hampshire,
Sussex, Essex, Herford, Buckingham, Bedford, Rutland, and perhaps Devon.
Outside of this area, he also noted some examples in Kent, Norfolk, Suffolk, and
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
96 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 97
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
98 Arriving at the Goal: Vowel Lengthening in Middle Germanic
with just these two consonants. When other consonants immediately following the
accented vowel, there is variation: doppelt (“double”), Koppel (“leash”), Krüppel
(“cripple”), Donner (“thunder”), Söller (“house addition supported by posts”)
versus Makel (“spot, stain”), Artikel (“article”). The lack of lengthening in words
of this type is often attributed to syncope of the unaccented vowel with cluster
formation. Be that as it may, it seems better to view these cases as original variation,
rather than attributing the short vowels to secondary re-shortening, which would
only serve to retain the status as a sound law for OSL.
Almost 100 years before Russ’ article, Paul (1884: 114–15), noted that there was
always lengthening if the consonant immediately following the accented vowel
was lenis: Leber (“liver”), Faden (“thread”), Magen (“stomach”); Ofen (“oven”),
Esel (“donkey”). When t and m are not followed by a sonorant, there is variation:
bitte (MHG bite; “please”), Gatte (MHG gate; “husband”), Kette (MHG keten(e);
“chain”), Sitte (MHG site; “custome”) versus Bote (“messenger”), Zote (“obscen-
ity, dirty joke”), Kröte (“toad”); Name (“name”), nehmen (“take”), schämen (“feel
ashamed of”), ziemen (“be proper”), Schemel (“stool”). In monosyllables with t, m,
there is no lengthening in words like the following: Blatt (MHG blat; “leaf”),
Schnitt (MHG snit; “cut, incision”), Tritt (MHG trit; “step, kick”). Russ (1969:
85–88) considers the absence of lengthening before t, m to be the rule and
lengthening to be the exception, so by his thinking t and m should be fortis
consonants (cf. d’Alquen 1979). He attributes the exceptions to analogy, spelling
variants, and dialect mixture, although, he admits, some words do not fit any of
these categories. The special status of t and m derives, for Russ, from the view that
consonant length is irrelevant in Middle High German, except for t/tt, m/mm
following short vowels. On the other hand, he considers strength irrelevant for r,
l, m, n. The overarching explanation is that consonant strength is primary, while
consonant length is secondary. Yet, given what we know about consonant strength
(cf. Section 1.2), it does not seem possible to tease out a distinction between
consonant strength and length in postvocalic consonants in Standard German.
Lengthening has taken place quite regularly in most Middle and Upper German
dialects. Looking first at Upper German dialects, we find that Swabian dialects
have regular lengthening, except before t, m in open syllables. Here are examples
from the dialect of the northern Black Forest (Baur 1967: 35–36, 55): ʃnaːb̥ l (Ger.
Schnabel; “beak, bill”), hoːb̥ lə (hobeln; “to plane”), faːd̥ ə (Faden; “thread”), b̥ oːd̥ ə
(Boden; “ground, soil”), waːɡ̊ ə (Wagen; “wagon, car”), fliːɡ̊ l (Flügel; “wing”);
haːfə (Hafen; “harbor”), oːfə (Ofen; “oven”), haːsə (Hase; “hare”), hoːsə (Hose;
“pants”); d̥ siːlə (zielen; “take aim”), ɡ̊ ʃd̥ oːlə pret. part (gestohlen; ‘stolen”), faːrə
(fahren; “haul, drive”), faːnə (Fahne; “flag”), hoːnex (Honig; “honey”), but fad̥ er
(Vater; “father”), b̥ od̥ (Bote; “messenger”), b̥ ɛd̥ ə (beten; “pray”), naːmə/name
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 99
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
100 Arriving at the Goal: Vowel Lengthening in Middle Germanic
“praise”), b̥ lɔːd̥ (MHG blat, Ger. Blatt; “leaf”), b̥ rɛːd̥ (MHG bret, Ger. Brett;
“board, plank”), d̥ ɔːx (Tag; “day”), wɛːx (Weg; “way”), hɔuf (Hof; “courtyard”),
ɡ̊ rɔːs (Gras; “grass”), viːl (viel; “much, many”), mɛːl (Mehl; “meal”), b̥ ɔːər (bar;
“bare”), loːm (lahm; “lame”), ʃoːm (Scham; “shame”), but ɡ̊ rob̥ (grob; “course,
rough”). Some short vowels are retained before endings with sonorants, espe-
cially -el, -er: ɡaβəl (Gabel; “fork”), ʃd̥ iβəl (Stiefel; “boot”), iβər (über; “over”),
fad̥ er (Vater; “father”), lɛrər (Leder; “leather”), oβə (oben; “above”), hamər
(Hammer; “hammer”), sumər (Sommer; “summer”). Rhotacism of old t is
a later change. Other Rhine Franconian dialects are similar, but a few descrip-
tions, like those of Mainz (Valentin 1934: 24, 32) and Saarbrücken (Kuntze 1932:
45), mention exceptions of the high vowels, especially u.
Short vowels in so-called open syllables are more widespread in Mosel
Franconian dialects. Many dialect descriptions mention exceptions of high
vowels, like those of Arel in Belgium (Bertrang 1921: 54ff), Saargau (Müller-
Wehingen 1930: 7–18), Saarhölzbach (Thies 1912: 38–44), and Ottweiler
(Scholl 1913: 23–36). For Sörth in Westerwald (Hommer 1915: 5–14), excep-
tions are reported for all vowels. Compare examples from the dialect of
Saarlouis (Lehnert 1926: 77), where most vowels are still short before endings
with sonorants: avæ (aber; “but, however”), huvəl (Hobel; “plane (tool)”), ivəl
(übel; “evil”), bød̥ əm (Boden; “ground, soil”), fad̥ æ (Faden; “thread”), hivəl
(Hügel; “hill”), ʃd̥ ivəl (Stiefel; “boot”), ʃβeəvəl (Schwefel; “sulfur”), mil
(Mühle; “mill”), hynix (Honig; “honey”), khinik (König; “king”). Old long
vowels are sometimes shortened in this dialect.
Ripuarian dialects show less vowel lengthening than Mosel Franconian dia-
lects, but it is not completely absent in any of them. Münch (1904: 30–35) denied
OSL for any Ripuarian dialect, but went on to give examples of lengthening
before fricatives and sonorants. In some dialects OSL is the rule, as in Aachen
(Welter 1938: 1–8), Schelsen (Greferath 1922: 6–13), and Eupen in Belgium
(Welter 1929: 22–23). In Malmedy and Vith (Hecker 1972: 71ff), on the other
hand, OSL is very limited. Compare, as one example, the following forms from
Schlebusch, now part of the city of Leverkusen, in Bergisches Land (Bubner
1935: 43), where there is some OSL of the mid and low vowels, but none of the
high vowels: æːvəns (eben; “even”), bəwæːjə (bewegen; “move”), kraːɣə
(Kragen; “collar”), waːɣə (Wagen; “wagon, car”), bɔːɣə (Bogen; “bow”),
jəlɔːɣə pret. part. (lügen “lie”), haːs (Hase; “hare”), hæːn (Hähne; “roosters”),
but zebə (sieben; “seven”), ʃtevəl (Stiefel; “boot”), øvə (über; “over”),
jevəl (Giebel; “gable”), jədrevə pret. part. (“drive, chase”), knɔdə
(Knoten; “knot”), flatə (Fladen; “flat cake”), ʃlofə (MLG slupen, Ger.
Pantoffel; “slipper”), spelə (spielen; “to play”), ʃtɛlə (stehlen; “steal”), hole
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 101
(holen; “fetch, get”), ʃɛlə (schälen; “to peel”), ʃamə (schämen; “feel ashamed
of”), nɛmə (nehmen; “take”), numə (Nummer; “number”).
The important question is whether the lack of lengthening in certain West
Middle German dialects is original or OSL has been reversed. These areas have
been documented in the handbooks and surveys (Schirmunski 1966: 181–87;
Wiesinger 1983a: 1092), but they point to a general lack of lengthening, which,
as we have seen, is not the case. The short vowels in these dialects cannot be
attributed to half-long vowels as the products of lengthening, which Victor
Michels (1921: 77–85) posited following Ten Brink (cf. Section 1.1).
In Ripuarian, the lack of OSL has been attributed to the Rhenish accentuation
by Schirmunski, but as we know from Swedish and Norwegian, tonal accents are
not antithetical to vowel lengthening in open syllables and OSL is not completely
lacking, by any means. The short vowels in open syllables in western Middle
German dialects are likely re-shortened, rather than unlengthened vowels, prob-
ably owing to the same type of analogy present in some words in English, i.e., to
short vowels in monosyllables. It seems highly unlikely that pockets in West
Middle German would be relic areas like those in southern Upper German or
northern and eastern Scandinavian, which are generally more archaic regions and
do not lie in the middle of Germanic territory.
3.2.3.3 Scandinavian
Open syllable lengthening is present in most of Scandinavian, aside from the
Swedish and Norwegian relic areas discussed in Section 3.2.1.1. Lengthening by
analogy in monosyllables was the rule, but is not consistent, especially in Danish,
where it is not widespread today. As mentioned earlier, MSL is often viewed as pre-
dating OSL in Scandinavian. The reasons for doubting this chronology were
presented above. The Scandinavian languages are outstanding for consonant
lengthening, which was a widespread alternative to vowel lengthening.
The geographic distribution of vowel and consonant lengthening, in what has
been called the “Great Quantity Shift” by Einar Haugen, is well documented in
handbooks of Scandinavian (Torp and Falk 1898: 17–25; Noreen 1903, 2: 118–19;
Noreen 1913: 141; Brøndum-Nielsen 1950: 379–81; Hansen 1962–71, 1: 388–89;
Wessén 1966: 24 and 1968: 90–93; Haugen 1976: 258–60; Pamp 1978: 28–29; see
also Hesselman 1901 and 1904). The following are the general contours of the
distribution: in Icelandic, Faroese, and Danish there was only vowel lengthening
and no consonant lengthening except for m in Danish. In Swedish and Norwegian,
vowel lengthening predominates in Göta Swedish and in West Norwegian, as in
West Germanic, while there is a tendency to consonant lengthening in Svea and
Norrland Swedish and in East and North Norwegian. Haugen saw a connection
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
102 Arriving at the Goal: Vowel Lengthening in Middle Germanic
between Danish or continental influence and the greater tendency to vowel length-
ening. Parts of South Swedish have been making a transition from the vowel
lengthening type to the consonant lengthening type. The main difference, however,
is between Danish, excluding Bornholm Danish, on the one hand, and the remain-
ing languages, on the other. While the majority of dialects have either long vowels
or long consonants in all accented syllables, Danish exhibits short monosyllables in
the present day.
Consonant lengthening in Scandinavian is much more widespread than in
German, the only other language where it occurred at the time of OSL. Aside
from m, which was also lengthened in parts of West Germanic, it was not only
Gmc. t, but the whole series of voiceless stops, which could lengthen. Gmc. p, t,
k are more common in Scandinavian than German, since there was no affrication
and spirantization via the Second Consonant Shift. Examples from Standard
Swedish, compared to Norwegian and Icelandic, will show the differing scope of
consonant lengthening: Sw. öppen (Nw. åpen, Ic. opinn; “open”), hoppas (Nw.
håpe, ON hopas; “to hope”), måtta (OSw. māt(t)a, Ic. máti, Da. maade; “must,
have to”), mycken (Nw. mye/megen, Ic. mikill; “much, large”), vecka (Nw. veka/
uke, Ic. vika; “week”); hammare (Nw. hammar, Ic. hamarr; “hammer”), himmel
(Nw. himmel, Ic. himinn; “heaven”), gammal (Nw. gammal, Ic. gamall; “old”).
Even from this limited set of data, it already becomes clear that Standard
Norwegian, based mostly on East Norwegian dialects, lies somewhere between
Swedish, based mostly on Svea dialects, and Icelandic.
In general, lengthening is more common before the voiced consonants, includ-
ing obstruents, like those in Sw. leva (OSw. liva/leva, OIc. lifa; “live”), häva
(OSw. hæfia; “heave”), bada (“bathe”), bade (“both”), mage (“stomach”), båge
(“bow”), than before voiceless obstruents in words of the type Sw. skapa (“to
shape, create”), dräpa (OSw. dræpa; “kill”), hata (“to hate”), prata (“talk”), äta
(OSw. æta; “eat”), baka (“bake”), kaka (“cake”). Lengthening before sonorants
except m is the rule: Sw. håla (“cave”), mala (“grind”). From all these examples,
it also becomes clear that the low vowels, ON a, æ, tended to lengthen more often
than the mid and high vowels, ON i, y, u, o, even in areas with a tendency to
consonant lengthening. By contrast, dialects of the vowel lengthening type have
long vowels where the other type lengthens the consonant, except when the
consonant is m. Compare examples from Västergötland (Götlind 1940–50, 3:
62): øːpən (öppen; “open”), drøːpə (droppe; “drop of liquid”), seːta (sätta; “place,
put”), heːta (hetta; “heat”), veːka (vecka; “week”), nøːkəl (nyckel; “key”), møːsə
(mosse; “peat-moss”), bøːra (borra; “bore, drill”), køːt (kött; “meat, flesh”), fløːt
(flott; “stylish”), beːk (beck; “pitch (liquid)”), but ʃepː (skepp; “ship”), ɡamməl
(gammal; “old”), kramma (krama; “hug”).
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 103
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
104 Arriving at the Goal: Vowel Lengthening in Middle Germanic
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005
3.2 Geographic Distribution of Vowel Lengthening 105
Downloaded from https://www.cambridge.org/core. University College London (UCL), on 08 May 2018 at 12:08:37, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781139540780.005