In this book, Kurt Goblirsch revisits and synthesizes many years of his previous publications with the worthwhile goal of establishing a common mechanism driving systemwide changes in Germanic obstruents. Like his earlier work, this book is notable for the tremendous amount of dialect data it reviews.

In the preface, the author rejects the “theoretische Richtung Noam Chomskys [Noam Chomsky's theoretical course]” (p. 5), and soon thereafter announces his alternative: “Besonders geeignet zur Bewertung der Lautverschiebungen ist der methodologische Ansatz des Prager Strukturalismus” (p. 14).¹

This is a curious stance, since neither “Chomskyan theory” (SPE?) nor “Prague School structuralism” is fundamentally a theory of sound change. Besides, Fourquet (1948) and others long ago pursued a Praguean approach to Germanic consonant shifts, and Goblirsch inherits the problems of those hoary analyses. For example, structuralism—just like early generative approaches—tends to be overly mechanistic, and ill-equipped for variation, something addressed directly in much modern work, like Ohalaesque historical phonetics, LabPhon, and Evolutionary Phonology (see Blevins 2004).

Goblirsch defines Lautverschiebungen [sound shifts] as “spontane, systematische Lautwandel, obwohl man sich auch darüber streiten kann, wie spontan genau jeder Lautwandel im Einzelnen ist. Sie sind insofern systematisch, als dass sie ein gewisses Teilsystem ändern” (p. 18).²

This definition could include many changes, but he posits precisely four shifts since Grimm's Law—the second (≠ High German!), Icelandic, Danish, and “partial” sound shifts—all unified by a restructuring of obstruent distinctions: a loss of phonological voicing is caused by the rise of aspiration (p. 236). He claims that all shifts follow a very precise path: voiceless stops aspirate, then aspiration gets lengthened and hardens to a fricative, creating an affricate. Its stop phase assimilates to the fricative phase. These steps represent “gradual obstruent weakening,” again following closely in Fourquet's footsteps.³

Goblirsch provides a typology of modern Germanic obstruent systems, arranged from the presumed most to the least archaic:

1. 1. The ASPIRATIONLESS TYPE, where the series are distinguished by occlusion, voicing, and length; exemplified by Dutch, West Frisian, southern Low German, and Scots. Goblirsch believes these never had aspiration, and have undergone no systematic obstruent changes since Grimm's Law.
2. 2. The COMMON TYPE, exploiting occlusion, voicing, aspiration, and length; Norwegian, English, and Low German. These have undergone “die Teilweise Verschiebung [partial shift],” having adopted aspiration.
3. 3. The VOICELESS TYPE, exploiting occlusion, aspiration, and length; eastern Danish, Icelandic, southwest Norwegian, and North Frisian. These have undergone Goblirsch's posited Second Consonant Shift but no “lenition,” abandoning voicing for aspiration, without any lenition.
4. 4. The LENGTH TYPE, exploiting only occlusion and length; Bavarian and Alemannic dialects.
5. 5. The ASPIRATED TYPE, exploiting only occlusion and aspiration; most of Danish, northern Low German, and some Central German dialects.

Thus, Goblirsch believes that Dutch and Scots represent archaic, practically West Germanic, patterns, while Upper German is quite innovative, and Icelandic takes an intermediate position. Widespread patterns of innovation/conservatism across Germanic—phonetic and phonological, but also lexical, morphological, and otherwise—suggest a different view: Icelandic and the southernmost German dialects are remarkably conservative almost across the board aside from the High German shift, while Dutch and English have evolved away from Northwest Germanic dramatically in virtually every regard. Some effort to harmonize these views would be most welcome.

As noted, Goblirsch imagines that all shifts happened in the sixth to eighth centuries, a chronological bottleneck that seems odd given the enduring pattern of similar changes stretching from the potentially pre-Grimm partial shift of double-stop clusters (+-pt(-) to -ft(-), etc.) to affrications underway today in Liverpool and Copenhagen fortis stops. A constant systemic pressure toward shift corresponds to the facts far better. His direct comparison to modern languages is also jarringly anachronistic. Icelandic was nonexistent at the time of his shift, with the settlement of Iceland, or landnám, only in the ninth to tenth centuries, yet he posits “eine isländische Lautverschiebung [an Icelandic sound shift]” centuries earlier. Many of the other systems he treats as archaic likewise did not exist in the eighth century and were then much later shaped by contact and areal pressures.

Where most scholars argue for aspiration from predialectal times, Goblirsch assumes that it arose in the sixth to eighth centuries (p. 74). As argued in these pages by Denton (1998), aspiration almost surely triggered North and West Germanic gemination, and so must have been present centuries earlier. Many varieties without aspiration surely lost it by language contact and others presumably by internal developments (a possibility Goblirsch implicitly dismisses, pp. 78–79). Positing the rise of aspiration independently in myriad dialects, as Goblirsch does, invites needless Occam's Razor concerns for little apparent gain.

In external history, Goblirsch sees Danish influence as driving this single shift across a vast territory, à la Zabrocki (1964, elsewhere). Little evidence is brought for this speculative proposal beyond an assumption that a few invaders or neighbors could transform the fundamental phonetics and phonology of large populations. A long tradition attributes Netherlandic and Scots voicing to historical language contact, yet Goblirsch rejects contact as the source for these phenomena, again without much explicit evidence or argument. Despite the importance of

the assertions made about language contact, no coherent view of language-contact change is presented; instead, it is evoked as deus ex machina.

Turning to sound change, Goblirsch's simple assumption (p. 47, and elsewhere) that aspiration directly “caused” the loss of voicing does not follow. Dutch obstruent cluster assimilations show systematic vestiges of an aspiration-language heritage, so that both features appear phonologically necessary (Iverson and Salmons 2003), and it seems generally prudent to acknowledge the possibility of featural overmarking.

One of Goblirsch's declared innovations is to identify the central role of consonant length in shifts (pp. 40–47). Krähenmann (2001, and elsewhere) provides masterful analyses of this issue for Thurgovian (and other Swiss dialects), covering history, phonetics, and phonology, yet her work is not cited at all. The study of gemination has also been an area of great progress in historical and synchronic phonology for years now, but that research too is mostly ignored.

Other basic questions—or outright mistakes—crop up, like these:

• [bull ] Goblirsch omits Yiddish from all lists of Germanic varieties without aspiration, perhaps over discomfort about the obvious role of language contact in its historical loss of aspiration (p. 22, and elsewhere).
• [bull ] He asserts that aspiration was “impossible” in s + stop clusters (p. 21). In fact, s-cluster aspiration is attested in various Indo-European languages (Greek, Armenian, Sanskrit).⁴

  See http://camba.ucsd.edu/phonoloblog/index.php/2005/11/17/saspirated-stop.

• [bull ] Goblirsch assumes frequently that aspiration characterizes INITIAL syllables, while the distribution is obviously tied to metrical structure or stress, not to word edges. (See the aspiration patterns in German [t], in fatal versus tolerant, for instance.)
• [bull ] Goblirsch denies a phonological contrast between OHG <t> from +d and <d> from +θ, arguing “vielleicht wollte man wenigstens den orthographischen Unterschied beibehalten” (p. 150).⁵

  “perhaps one wanted to at least maintain the orthographic distinction.”

  This surely oversteps the facts, given how many dialects maintain a distinction down to the present day.

The references are vast, taking up pages 246–301 set in small type, and the book is largely built around earlier work, mostly criticizing it. Accordingly, one would expect that the volume would have great value as a review of the literature, but unfortunately the Forschungsbericht overlooks much important recent work, as already exemplified several times above. Other gaps include most of the modern body of work on laryngeal features (like fortis/lenis and spread glottis), some work on consonants in German and Germanic dialects, and so on. For example, among the missing names that one might hope to find on phonetic/phonological features relevant to Germanic obstruents are William Barry, Andrew Butcher, Sean Fulop, Gunnar Ólafr Hansson, Pétur Helgason, Manfred Pützer, Catherine Ringen, Philip Spälti, Bert Vaux, and Urs Willi. The research review also misses the spirit of some work, including my own. Even so, much of the literature review is superfluous for specialists. For example, the treatment of IE obstruents could be truncated by simply accepting the now generally acknowledged view that “voiced aspirates” were murmured.

Coverage in the name index is spotty—not including many scholars whose work is referred to—and the subject index is sometimes wrong, for example, the entry for binnendeutsche Konsonantenschwächung. Typographical errors occur especially in names—Wissman (p. 32, but Wißmann in the references), Hans Heinrich Hock (lege Henrich, p. 67)— and in the references (where the alphabetization is also at times incorrect)—“Imengard” for Irmengard (p. 260), “lanrygeal” for laryngeal (p. 266), etc.

Summing up, Goblirsch struggles mightily to make this revival of Fourquet and Zabrocki work, but he faces insurmountable challenges. The proposed chronology is unworkable, the assumed role of language contact implausible, and the understanding of sounds and sound change outdated and problematic. By the end of the book, one fears that the theoretical salvo at the outset was meant as a license to ignore much work from the last half century. That is a shame.