Why this is the new book on German phonology that we needed

Virtually every publication on the phonology of Standard German – that is, the standard variety of German spoken in Germany – references Richard Wiese’s (Reference Wiese1996) The Phonology of German, for at least two reasons. First and foremost, it is an excellent book. Second, however, it is essentially the only English-language volume that provides a comprehensive account of the phonology of German. This in turn means that beyond its many undoubted qualities, it has had a three-decade monopoly as the default reference on anything regarding the synchronic structure of the German sound system. Monopolies are rarely all too beneficial, though, and even the greatest works can be subject to the ravages of time, at least regarding certain aspects – after all, not only languages change, but theories do, too, such that the importance of specific analytical problems and perspectives can be in flux over time. Particularly against that backdrop, Caroline Féry’s new book-length treatment of the German sound system is timely and more than welcome.

Ambitious in its scope and equipped with machinery that utilizes various advances in phonological theory over the past few decades, the book is a major addition to the study of German phonology, with the potential to be the standard work for decades to come. In its comprehensive coverage of data and in its analytical depth, it is at any point at least comparable to Wiese’s Phonology of German, and certainly extends beyond it in the realm of prosody, that is, the structures located above individual segments. Therefore, the present book can be expected to take the place at the very top of book-length treatments of German phonology.

Who is the right audience for this book?

The intended target audience for the book is “advanced students of both linguistics and German, as well as individual scholars seeking a one-stop resource on the topic,” which is a sensible categorization, certainly with regard to linguistics students. While being detailed, thorough, and carefully building up its content, this volume would probably not be the perfect choice as a textbook for a first undergraduate introduction to phonology, as its contents might still be a little too heavy for the uninitiated. It will be a great source for advanced linguistics students, though, and nothing less than a must-read for any senior linguist interested in the study of German or of Germanic languages more broadly. As such, this volume could certainly serve as a textbook in advanced linguistics classes on the phonology of German, and at least as a supplementary reading in graduate courses on phonological theory.

To what degree it will be accessible to “advanced students of German” may depend on an individual’s prior knowledge of (generative) phonology; for German students that are interested in Féry’s book but feel like they may not yet have enough background knowledge, I recommend reading a more basic introduction first, such as, for example, Hall’s (Reference Hall2011) German-language phonology textbook, which should do the job of preparing them for Féry’s volume just fine.

Structure of the book

The book is sensibly structured into ten chapters, moving from the very atoms of phonology to the highest prosodic units at the phrase level. We find thorough discussions of segments and segmental features, followed by a treatment of moras, syllables, and feet, all of which is supplemented by insightful analyses of segmental processes and prosodic properties, such as an in-depth study of the German stress system, which illustrates and elaborates on Féry’s approaches to word stress over the past 30 years or so. The detailed treatment of phrasal phonology is particularly laudable, with many things to learn for even the most advanced scholars of German phonology.

Since the analyses are thoroughly formalized in Optimality Theory (OT), it also helps that at the beginning of the book, we find a concise but useful introduction to basic principles of OT, at this point still the most widely used framework in phonology (as such, understanding at least its basic principles is recommended for anyone studying sound systems). Theoretically oriented readers will certainly appreciate Chapter 10, which carefully shows how the OT analyses of individual phenomena relate to one another, insightfully discussed in the main text and illustrated with so-called Hasse diagrams that show dominance relations between constraints. This is a prime example of how to present the overarching takeaways of an OT analysis, a step that is too often omitted in scholarly works.

In doing so, the final chapter also illuminates how the analyses of different phenomena build on and relate to each other in sensible ways, leading to the kind of analytical coherence that is exactly what one would ideally hope for in any book on the phonology of a given language. To give an example, Hasse diagram (4) on p. 430 concisely illustrates the commonalities in the distribution of [ŋ], [h], the glottal stop and glides in formal terms, demonstrating how crucial distributional similarities of these segments are directly reflected in the formalization via shared constraints rankings.

Is the phonology of German now “solved”?

Given my well-deserved high praise of this book, a reader may now wonder whether we might conclude that the phonology of (Standard) German is essentially solved, and that everyone else can pack up and focus on studying other languages. Well … maybe not quite. One thing that is inherent in even the most impressive formal treatments is that any scholar has to make analytical decisions, and even if everything falls into place, other scholars might disagree with certain assumptions and prefer to take a different route. For that reason, even in an approach as coherent as Féry’s, there will necessarily be aspects to debate and reevaluate.

As an example, let us consider certain details of the analysis of (alleged) similarities between diminutive umlaut and the place of dorsal fricatives. For space constraints, my reasoning below may be easiest to follow for more specialized readers, as it presumes knowledge of certain theoretical concepts and contains some more technical language than the rest of this review. Everyone not interested in such details right now should feel free to skip to the last paragraph of this review.

That being said, consider first the well-established observation that the dorsal fricative after a front vowel will typically be realized as a palatal [ç] in German (1a), while it will be a velar [x] after a back vowel (1b); after consonants, it will be a palatal [ç] (1c). There exist additional restrictions and generalizations, but these are not relevant for our point here. Moving forward, I will refer to the interactions in (1) as the basic pattern.

In a nutshell, Féry assumes that dorsal fricatives are underspecified for backness; underspecification in Féry’s analysis means that the respective fricatives do not carry a backness feature in their lexical representation, but that the feature is inserted by the grammar. After consonants, the front feature [-back] is inserted due to a default preference, resulting in the palatal fricative [ç] with the feature [-back]; the velar fricative [x] has the feature [+back]. After vowels, the constraint in (2) determines the distribution:

The effects of DFA are such that after back vowels, a velar fricative will be preferred, such that the vowel and the fricative will share the feature [+back]. After front vowels, DFA will prefer a palatal fricative, with the segments sharing [-back]. The default preference for [ç] and the effects of DFA thus correctly derive the distribution of dorsal fricatives in German. While this specific analysis is not particularly surprising or controversial, Féry extends the same general approach to umlaut triggered by the diminutive suffix -chen, which always starts with a [ç].

Let us look at the main facts first: As we can see in (3), back vowels in base forms preceding the diminutive suffix are typically umlauted to front vowels. In Féry’s analysis, it is the fricative that determines the vowel quality of the preceding consonant, the opposite of the basic pattern where the vowel determines the place of the following fricative. Féry derives the diminutive facts from the same general principles as the basic pattern, again utilizing the DFA constraint in (2). The crucial ingredient for her analysis is that the palatal fricative [ç] is specified for [-back] in diminutives, rather than being unspecified, as we find it in the basic pattern. This has consequences for the relevant interactions: While vowel backness influences underspecified dorsal fricatives, the specified palatal fricative [ç] determines vowel backness in diminutives. This is possible because the constraint DFA can be satisfied in both directions; that is, it does not matter what the trigger and the target are, as long as the vowel and the dorsal fricative agree in place.

The idea to use the same DFA constraint plus certain principles of underspecification to account for both phenomena with the same general machinery is as novel as it is ingenious. One remaining question is whether it works with the established formal apparatus, and to be honest, I am not entirely sure that it does.

Consider the definition of DFA in (2) again, which states that “a back vowel and a following dorsal fricative share the same specification for [back].” This definition does not explicitly specify whether the interaction must be strictly local or not, that is, whether a segment can intervene between the vowel and the fricative. In phonology, the default assumption for vowel–consonant interactions would probably be that the segments have to be directly adjacent, in which case the definition in (2) would be fine; as we will see, however, the empirical picture for German is more complicated than that.

First of all, note that, as shown in (4a), the local definition of DFA works well for the basic pattern triggered by vowels, where an intervening consonant blocks the realization of the dorsal fricative as [x] after a back [a] in [mançə] ‘some’. Interestingly, however, the local interpretation of the constraint does not work for diminutive umlaut. As we observe in (3), segments intervening between the palatal fricative and the preceding vowel can in principle be skipped in umlaut triggered by diminutives. If we use a local definition of DFA, we therefore predict an incorrect outcome for the diminutive patterns. This is exemplified in (4b), where [a] is fronted to [ɛ] despite there being a consonant [n] intervening between the vowel and the palatal fricative, which should block the process as per the local definition of DFA. As such, the local interpretation of DFA only derives the basic pattern correctly, but not the diminutive pattern.

Conversely, the non-local interpretation of DFA makes correct predictions only for diminutive umlaut triggered by [ç], as shown in (5b). For the basic pattern triggered by vowels shown in (5a), however, the non-local definition incorrectly predicts that the back vowel [a] should trigger a [+back] realization of the dorsal fricative, favoring unattested *[manxə]. To make the current analysis work, we would thus need to interpret the constraint in different ways depending on which phenomenon we are currently accounting for. This is problematic, but this appears to be the way it is done in the book.

In my view, the most reasonable conclusion to draw here would be that the agreement analysis may not be the right way to account for umlaut in diminutives. In addition to the DFA issue, another drawback of analyzing diminutives that way is that it is hard to extend this approach to other diminutive suffixes triggering vowel fronting in similar ways, such as the standard form -lein or dialectal -ke and -ge, where a feature sharing analysis along the lines of -chen is not available (Féry herself acknowledges the issue for -lein in footnote 34). Furthermore, the analysis relies on assumptions about foot structure that may deserve further discussion, but treating this issue here in more detail would go beyond the scope of this review.

In the grand scheme of things, however, having certain issues left to debate is only good and helpful to move our field forward. All in all, Féry’s book does a tremendous job of providing a rich and comprehensive source of all the most important things one could think about when studying the sound system of German. For all the reasons mentioned throughout this review, I recommend the lecture of this volume in the strongest possible terms.