Middle English verbs of emotion and impersonal constructions: Verb meaning and syntax in diachrony, a revised version of Ayumi Miura's PhD thesis (Manchester University, 2011), focuses on a very well-studied area in English historical syntax and semantics, impersonal verbs and constructions. The book is the fifth volume to be published in the series Oxford Studies in the History of English, which in fact opened with another monograph on this topic (the monumental study on Old English by Möhlig-Falke Reference Möhlig-Falke2012). This is clear proof that interest in impersonals is still very much alive, despite being on the front line of diachronic research since the late nineteenth century. In recent years linguists have applied new methods and recent theoretical frameworks to this well-explored area and have still been able to find new niches for research. This is the case with the present book: Miura's study has a lexico-semantic focus and addresses a very specific semantic domain in which impersonals are especially frequent, that of verbs of emotion, looking at the Middle English period in particular. Motivated initially by Denison's (Reference Denison, Adamson, Law, Vincent and Wright1990) article, Miura carries out a comprehensive study of both impersonal verbs and their ‘outer perimeter’ (Denison Reference Denison, Adamson, Law, Vincent and Wright1990: 122), paying attention not only to impersonal verbs themselves but also to near-synonyms which have not been attested in the impersonal construction in the historical record. This novel approach allows her to pursue the main goal of her research, namely to discover the intricate factors which determine the occurrence of individual verbs in the impersonal construction and to shed light on impersonal/non-impersonal variation.
The volume is organized into six chapters: chapters 1–3 present the theoretical and methodological background, while chapters 4 and 5 describe Miura's empirical study. The main conclusions are summarized in chapter 6, after which an index of concepts and verbs is provided.
Any study on impersonal verbs and constructions must begin by clarifying how the terms ‘impersonal construction’ and ‘impersonal verb’ are understood (see Méndez-Naya & López-Couso Reference Méndez-Naya and López-Couso1997 for problems of terminology here). As we know, the central characteristics of impersonal constructions are their lack of a nominative subject and the presence of a verb inflected for the third-person singular. Miura starts from the basic complementation patterns listed by Möhlig-Falke (Reference Möhlig-Falke2012: 6–8): (A) no complement at all (e.g. norþan sniwde ‘it snowed from the north’); (B.a) a single complement realized by a finite clause (e.g. þa gelamp þæt-clause ‘then it happened that-clause’); (B.b) a single complement realized by an accusative/dative NP of person (e.g. me acc/dat hyngrede ‘I was hungry’); (C) two complements, an accusative/dative NP of person and a second complement, which can be phrasal (NP or PP) or clausal (finite or non-finite) (e.g. ðone geðyldegestan acc scamað ðæs siges gen ‘the most patient is ashamed of the victory’, for an example of a phrasal complement). She decides to limit the term ‘impersonal construction’ to those patterns which show an overt oblique Experiencer, namely (B.b) and (C) (p. 6). The rationale for this restriction is that only in such patterns is variation with a nominative Experiencer available, and it is this variation that is crucial when addressing her research aims. ‘Impersonal verbs’ in this study are those ‘which are known in the literature to have appeared in impersonal constructions, even though they may also have been employed in other syntactic patterns’ (p. 7), while ‘non-impersonal’ verbs are those which are not known to have been used impersonally. This definition of impersonal verbs can sometimes be problematic, in that Miura will indeed find impersonal uses of verbs which had been considered non-impersonal (see, e.g., p. 201 on verbs of Mental pain/suffering). In addition to terminological issues, chapter 1 also sets out the aims of the study (see above) and justifies her decision to focus on a particular semantic domain (that of emotion, ‘the largest semantic class that allowed occurrence in impersonal constructions in Old and Middle English’, p. 19) and on a particular historical stage (the Middle English period, less widely studied than Old English, p. 16).
One of the strengths of this study is undoubtedly its methodological care; the reader is continually told what the author has done, which decisions she has made, and how particular examples have been analysed and tallied. Chapter 2 is devoted to theoretical and methodological considerations. Here the old question of the bad data in historical studies is addressed, as well as more specific issues like the interpretation of negative evidence and the significance of nonce examples. According to Miura, in the case of impersonals the implications of negative evidence may vary depending on the frequency of the verb: if a very frequent verb is never used impersonally we are safe to take negative evidence at face value; if it is a rare verb (attested less than ten times in the material), non-occurrence in the impersonal construction may be accidental (p. 23). This clarification is relevant because Miura will resort recurrently to negative evidence in the data analysis. In this chapter the author also presents the methodology and findings of a number of studies on the syntax and semantics of near-synonymous verbs as well as research on psych-verbs in modern languages. This allows her to identify a number of potentially relevant factors for variation which she will test later on in her detailed analysis of impersonal and non-impersonal verbs of emotion (chapter 5).
Methodological issues are also dealt with in chapter 3, where Miura justifies the selection of this particular group of verbs, and presents the Historical Thesaurus of the Oxford English Dictionary (HTOED), which she uses to categorize the semantic field of Emotion. Here the author explains the classificatory system of the HTOED, as well as its advantages and disadvantages. Chapter 3 is very short (pp. 47–55) and it would perhaps have been better to integrate it into the preceding chapter, which also has a clear methodological orientation. Similarly, the process of list-compiling, thoroughly described in section 4.1 (chapter 4), and the justification for the use of the entries in the Middle English Dictionary (MED) as a database (section 5.1 in chapter 5) could have been accommodated in chapter 2.
Based on the categorization of Emotion in the HTOED and checking her results with Möhlig-Falke's (Reference Möhlig-Falke2012) data, Miura identifies seven semantic categories which include impersonal verbs from the Old English period until 1500 (the approximate date of the demise of the impersonal construction): (i) Pleasure/enjoyment; (ii) Mental pain/suffering; (iii) Anger; (iv) Hatred/enmity; (v) Pity/compassion; (vi) Humility; and (vii) Fear. These seven categories are thoroughly analysed in chapters 4 and 5, which constitute the core of Miura's study. There the author makes use of lexicographical evidence in two different, but complementary, ways: to explore the semantics of impersonal verbs of emotion through the dictionary definitions in Bosworth & Toller (BT(S)), the Dictionary of Old English (DOE) and the MED (chapter 4), and to investigate the semantic/syntactic distinctions between impersonal and non-impersonal verbs of emotion by means of an exhaustive analysis of the quotations of the relevant MED entries (chapter 5).
The focus of chapter 4 is exclusively on impersonal verbs; it provides the dating of the impersonal use for each semantic category as well as attempting to reach semantic generalizations across categories. Miura shows that most of the semantic classes mentioned above already contained impersonal verbs in the Old English period (namely Pleasure/enjoyment, Mental pain/suffering, Hatred/enmity, Pity, Humility). With verbs of Anger and Fear, by contrast, impersonal use is an early Middle English innovation, probably originating in the West Midlands, which seems to be crucially related to the presence of some sense of stimulus (shuddering in the case of Fear and irritation with verbs of Anger), a constraint which seems to weaken after the fourteenth century. By studying dictionary definitions she detects a clear correlation between causative meaning and impersonal use (precisely one of the conditioning factors for psych-verbs in modern languages that she introduces in section 2.4.1), as well as a tendency for impersonal verbs of emotion in most semantic classes to be stative as far as aspect is concerned.
In chapter 5 the focus shifts to include the outer perimeter, that is, those verbs of emotion which have not been attested in impersonal usage. The time span studied is restricted to the Middle English period, which is illustrated by data drawn primarily from the quotations in the MED entries of the studied verbs. In this chapter the author pays special attention to the syntactic and semantic factors which had proved relevant for near-synonymous verbs in early English and psych-verbs, namely: (i) constructional patterns, i.e. do differences between impersonal and non-impersonal verbs have a correlate with the constructions in which they appear? (ii) Target of Emotion (ToE), i.e. do impersonal and non-impersonal verbs differ as to the animacy of their ToE? (iii) argument alternations, i.e. are any alternations (e.g. the conative alternation) available only with impersonal or with non-impersonal verbs? (iv) causation, i.e. do non-impersonal verbs have any causative meaning or show a syntactically causative use? and (v) aspect, i.e. are there any aspectual differences between impersonal verbs and non-impersonal verbs? Miura sets out to check for all these factors in each of the seven semantic classes. For the sake of comparison, she also includes in her study three further HTOED emotion categories which did not yield any impersonal verbs (Jealousy/envy, Pride and Courage). In order to gauge the degree of semantic similarity between impersonal and non-impersonal verbs, she looks at their occurrence in word pairs (which would point at the existence of slight semantic or stylistic differences), and as manuscript variants (which argues for semantic equivalence). Co-occurrence with different kinds of adverbials is also used as evidence to identify subtle semantic differences. In turn, aspectual differences are tested by studying the co-occurrence of verbs with eventive adverbials and the habitual/non-habitual interpretation of the simple present.
The careful analysis reported in chapter 5 enables Miura to identify a number of interesting trends, which she presents in section 5.11 and in chapter 6:
(i) Most impersonal verbs of emotion share causative meanings. The connection between causation and impersonal usage is also confirmed by the lack of causative meaning with verbs of Jealousy/envy and Pride, which were never attested in the impersonal construction. However, causation is not a necessary condition for impersonal use, since there are causative verbs which do not occur in the impersonal construction, and, contrariwise, impersonal verbs exist which seem to lack a causative meaning.
(ii) Impersonal verbs in most emotion categories tend to have a stative reading, thus allowing a non-habitual interpretation in the present tense and avoiding collocation with eventive adverbials. Stativity is particularly salient when impersonal verbs occur in the impersonal construction. In this respect, Miura's findings for Middle English cast doubt on Möhlig-Falke's (Reference Möhlig-Falke2012: 86) association of impersonal verbs of emotion with dynamic events. As was the case with causation, aspect is not found to be the only determinant of impersonal use.
Interestingly, the features of causativity and stativity establish a clear link between impersonal verbs of emotion and psych-verbs.
(iii) Constructional patterns: In some categories, impersonal and non-impersonal verbs are distinguished by the (non-)availability of the Experiencer-subject passive construction (e.g. God was worst apaied with her thanne before ‘God was most displeased with her than before’), with impersonal verbs lacking such a pattern. This is in line with Möhlig-Falke's (Reference Möhlig-Falke2012: 193–7) finding that Old and Middle English impersonal verbs are rarely attested in the passive. Furthermore, the availability of reflexive constructions sometimes distinguishes impersonal from non-impersonal verbs.
(iv) Impersonal verbs in some of the emotion categories (namely Hatred/enmity, Pleasure/enjoyment, and Pity/compassion) participate in the conative alternation. For example (p. 170), the impersonal verb loath can occur in both the transitive and the intransitive Experiencer-subject constructions, with its ToE realized by an accusative NP (e.g. þat me is lef, all acc she loþes ‘she hates all that is pleasing to me’) or by a PP ( of hir life , she gan to loothe ‘she was disgusted with her life’). The non-impersonal verb hate, by contrast, is limited to the transitive variant (e.g. hir lif acc sche hateþ ‘she hates her life’). This suggests that ‘verbs with high transitivity cannot in principle be impersonal’ (p. 230).
Behaviour as regards constructional patterns and the conative alternation seems to indicate that Middle English impersonal verbs of emotion form a coherent set featuring low transitivity. This also ties in well with their correlation with stativity, since stativity is itself one of the indicators of low transitivity according to Hopper & Thompson (Reference Hopper and Thompson1980).
(v) Animacy of the ToE distinguishes impersonal from non-impersonal verbs in only some of the semantic categories, with inanimate ToEs correlating with impersonal use especially before the fourteenth century (see also Möhlig-Falke's similar findings for Old English, 2012: 109, 150), but the data seem to indicate that animacy cannot be regarded as a critical determinant.
(vi) Contrary to what has been proposed in earlier studies (e.g. McCawley Reference McCawley, Steever, Walker and Mufwene1976, or more recently Möhlig-Falke Reference Möhlig-Falke2012), Miura's data for Middle English verbs of emotion fail to show a strong correlation between impersonal use and immediate/uncontrollable feelings. Interestingly, however, while impersonal verbs can indeed have the ability to describe immediate feelings, non-impersonal verbs like love and hate conspicuously lack it. This coheres with psychological definitions of ‘emotions’, which are characterized as brief and episodic vs ‘affective commitments’ like love, hate or jealously, which tend to last longer (see Diller Reference Diller2008). Curiously enough, these psychological definitions seem to have a counterpart in syntax, since verbs denoting these long-term feelings (notably love and hate) proved to be incompatible with impersonal use in Middle English.
After her meticulous analysis of the data, Miura concludes that none of the factors identified in her study can be taken to be the sole determinant of occurrence in the impersonal construction. Rather, it is when we find this ‘constellation of properties’ (p. 238) that the impersonal use is favoured. Thus, it can be predicted that a Middle English verb of emotion is prone to occur in the impersonal construction if it is causative, ranks low in transitivity, tends to occur with inanimate ToEs and is able to express episodic or immediate mental states in certain contexts. By contrast, if a verb lacks causative uses, ranks high in transitivity, shows a preference for animate ToEs and generally expresses long-term feelings, it is not likely to occur in impersonal constructions. However, as Miura acknowledges, ‘none of these parameters is rigid’ (p. 238).
Ayumi Miura's monograph constitutes a most welcome addition to the abundant literature on early English impersonals. The comprehensive study of emotion verbs, both impersonal and non-impersonal, allows her to disentangle the ‘complicated impersonal/non-impersonal borderlines’ (p. 238), which, as she shows, are sometimes rather fuzzy, and become even fuzzier as the Middle English period advances. In her approach, it is the semantic features of the verb that trigger its use in the impersonal construction, although she sometimes seems to argue that occurrence in the impersonal construction prompts a particular interpretation of the verb as would be suggested from a constructional point of view (e.g. the uncontrollable feeling interpretation of some verbs of Hatred/enmity seems to be triggered only when they occur in the impersonal construction, p. 173). The author's analyses and interpretation of the individual examples are very sound, even though at times this reader does not completely agree with them (e.g. example 4.66, p. 84 as an impersonal construction) or is not fully convinced of the semantic nuances the author strives to find. The presentation of the philological material is extremely careful: Miura provides over 400 illustrative examples, all of them fully glossed and translated, conveniently contextualized so that the reader may also reach his/her own conclusions. I particularly enjoyed the detailed discussion of alternative analyses (e.g. examples 5.227 and 5.228, p. 202), her interpretation sometimes differing from those of her lexicographical sources, or providing more nuanced accounts. It is worth noting that some of the examples quoted in the book come from the author's own readings of Middle English texts, which shows her familiarity with the Middle English sources in the best philological tradition. Also praiseworthy is the extensive bibliography, notably including psychology literature for the study of emotions. I have already mentioned her methodological care and honesty (never leaving aside evidence or factors which may be problematic), which is indeed one of the highlights of the present book, as is her innovative use of the HTOED and the MED quotation database as a framework for a syntactic and semantic study of early English verb classes. One of the very few caveats is that the book seems to retain occasional traces of its origins as a PhD thesis, such as explanations whose relevance to the topic are not immediately obvious (e.g. the description of Carroll's (Reference Carroll1997) study on Middle English verbs of cooking (p. 33ff.), used to support Levin's (Reference Levin1993) hypothesis that the syntactic characteristics of a verb are semantically determined is perhaps too detailed), and the overabundance of endnotes.
Overall, Ayumi Miura's Middle English verbs of emotion and impersonal constructions: Verb meaning and syntax in diachrony constitutes an important contribution to our knowledge of impersonals in English, and will therefore be of relevance to those working on Early English syntax. In addition, it will undoubtedly be also of interest to scholars working on diachronic changes in argumentation structure, to linguists concerned with lexical semantics and to lexicographers, since the study unveils semantic nuances which thus far have escaped dictionary makers. The study also shows that even topics which have been as extensively researched as Early English impersonals can be found to offer new perspectives in light of recent theoretical approaches and methodologies.
