Abstract
The Acquisition of Derivational Morphology: A cross-linguistic perspective edited by Veronica Mattes, Sabine Sommer-Lolei, Katharina Korecky-Kröll and Wolfgang U. Dressler (2021) is the latest edited volume to emerge from the International Cross-Linguistic Project on Pre- and Protomorphology. Previous volumes have focused on the acquisition of verb inflection (Bittner et al., 2003a), nominal inflection (Stephany & Voeikova, 2009), diminutives (Savickienė & Dressler, 2007), semantics and morphology of adjectives (Tribushinina et al., 2015) and nominal compounds (Dressler et al., 2017). The protomorphology research agenda has consistently highlighted the importance of a cross-linguistic perspective, arguing this is the only approach that allows us to determine which observations can be explained by universal processing mechanisms and which are explained by typological characteristics of the target language. This latest volume contains analysis of language data from 10 languages from 3 language families, as shown in Table 1. Commitment to cross-linguistic study is undoubtedly necessary to help progress our understanding of first-language acquisition. However, it should also be noted that the language sample considered in this volume only draws on languages from three language families and that the sample is heavily biased towards Indo-European languages. This bias towards certain types of languages and cultures represents an ongoing challenge to the field of language acquisition (e.g. Kelly et al., 2015; Kidd & Garcia, 2022; Pye, 2021) and is acknowledged as a limitation by Mattes and Dressler (2021, p. 300) at the conclusion of the volume. See recent commentary in First Language on Kidd and Garcia (2022) for further discussion of this issue.
Language families, branches and languages.
Source: Adapted from Dressler et al. (2021, p. 10).
The volume is constructed similarly to past volumes in the series. The Introduction (Dressler et al., 2021) lays out the theoretical basis of the protomorphology research programme and outlines questions to be considered. The subsequent chapters, by a variety of authors (Chapters 2–11), dutifully address these questions for each language before the threads are tied together in a final Conclusions chapter (Mattes & Dressler, 2021). The volume aims to describe the emergence and early developmental patterns in derivational morphology in each of the languages considered. Each analysis chapter examines longitudinal dyadic interactional data between children, most typically before age 3, and a caregiver. Each chapter provides an initial overview of the derivational sub-systems of the language being considered divided into the categories noun, verb and adjective. They then proceed to describe the prevalence of derivational morphology in children’s speech for each of the sub-systems and compare this with child-directed speech (CDS) from the same interaction. This approach to analysis is argued to be key to determining the relationship between input and output. The French chapter (Kilani-Schoch & Xanthos, 2021) also includes a comparison with adult-directed speech.
Protomorphology divides early morphological development into three phases: (1) premorphology, (2) protomorphology and (3) morphology proper. At the stage of premorphology, children produce rote-learned forms and are unaware of the morphological system. The protomorphological phase starts when children begin to detect morphology as a way of decomposing and composing meaning. The final phase, morphology proper, is a logical endpoint where children have acquired adult-like morphology. Practitioners of protomorphology regularly use the criterion of a mini-paradigm (Bittner et al., 2003b) to determine when a child has reached the protomorphological stage. The proposed definition adapted for derivational morphology is
that a child has to produce spontaneously at least three derivatives with each lexical base, which itself occurs twice as frequently as the autonomous word or, more importantly, as part of other derivatives or compounds in the same or adjacent recordings, neither in formulaic nor in immediately imitating speech and that the same affix or non-concatenative process recurs on two more occasions. (Dressler et al., 2021, p. 8)
Mattes and Dressler (2021, p. 291) acknowledge that meeting this criterion does not necessarily indicate that a child has established a morphological pattern or rule. However, they argue it is a useful data point that identifies the stage at which a child has the information available to them to detect a morphological pattern, described as potential productivity. For those authors who utilised this criterion, the first mini-paradigms for derivation were fulfilled at the ages shown in Table 2 (Mattes and Dressler, 2021, p. 291). No mini-paradigms were identified in the Danish data before age 3;0.
Age of first derivational mini-paradigm by language.
This measure alone does not provide a particularly useful point of comparison, as it is unclear what part of the derivational system this potential productivity is observed in, and it does not identify the relevant factors that cause a derivational pattern to be acquired earlier or later. Across the languages and derivational sub-systems analysed, the factors identified as influencing acquisition typically were the frequency of a particular derivational pattern in CDS, alongside the morphotactic and morphosemantic transparency of that pattern. Greater frequency and transparency in general led to earlier usage of derivational forms and the earlier emergence of mini-paradigms. While throughout the volume the various authors provide detailed description of the language data, it would benefit the reader to be able to draw comparisons between the different derivational subsystems across the various languages. This could be achieved through the quantification of factors such as transparency, type/token frequency and the overall ‘productivity’ of a derivational subsystem in terms of the attested members of a derivational pattern, which could then be related to the mini-paradigm criterion.
It should also be noted that although a morphological pattern may be potentially productive, as evidenced by the identification of a mini-paradigm in a child’s speech, this does not dictate that a child must eventually identify and utilise this pattern. For example, in the polysynthetic Australian language Murrinhpatha (e.g. Street, 1987; Walsh, 2011), children acquiring bipartite stem verbs constructed of interchangeable lexical and classifier stems which together account for verbal semantics and argument structure (Nordlinger & Caudal, 2012) do not readily divide meaning across these elements. Although many stem combinations are transparent, there are also many opaque combinations (Forshaw, 2021, pp. 18–25). The lack of morphosemantic transparency and productivity across the bipartite stem verb system results in children showing little evidence of needing to decompose and compose meaning from individual stem elements before age 6 despite using a range of diverse combinations (Forshaw, 2021, pp. 137–138). When reading this volume, it is important to keep in mind that references to productivity throughout it are often references to potential productivity. A child may have sufficient evidence to detect a morphological pattern, but this does not necessarily mean that they have detected one and are able to utilise it.
There is some discussion of neologisms throughout the volume which provide the clearest evidence that children have acquired a derivational pattern. However, these types of errors are typically rare in longitudinal data, meaning there is a need to rely on other criteria for determining the acquisition of productivity. It is also noted that, given the age range considered, these may emerge later for many of the derivational systems investigated. Indeed, it would be informative to examine children’s development beyond age 3, in particular regarding the acquisition of less frequent and transparent derivational patterns which are largely frozen forms in the age range considered here.
Pragmatics was also argued to be a relevant factor in the acquisition of some derivational patterns, with patterns useful to a child’s communicative goals being acquired earlier (e.g. Sommer-Lolei et al., 2021, p. 133). The role of pragmatics in the acquisition of derivational morphology and morphology in general is an interesting area for future investigation given that the way in which language is used to do things has been shown to affect acquisition of particular language structures in some languages. For example, pragmatics has been argued to influence the acquisition of passive constructions in Sesotho (Demuth, 1989), the ergative marker in Samoan (Ochs, 1988) and imperative constructions in Finnish (Laalo, 2003) and Murrinhpatha (Forshaw, 2021, Chapters 5–6).
This volume provides a valuable contribution to the field of language acquisition and will be of interest to those actively researching the acquisition of morphology and the cross-linguistic study of language acquisition more generally. It is also a great example of what can be achieved through ongoing cooperation and collaboration among researchers working to address common questions. This cooperative endeavour would be strengthened in future by quantifying relevant typological factors to allow for better crosslinguistic comparison and by considering a set of languages more representative of humankind.
Footnotes
Acknowledgements
Thanks to Barbara Kelly for helpful feedback on this review.