With their edited collection Phraseology and Style in Subgenres of the Novel, Iva Novakova and Kirk Siepmann introduce an interdisciplinary approach to genre studies, at the intersection of corpus linguistics, computational linguistics, and stylistics. The primary concern is to explore the recurrent features of fictional genres and their general functions, focusing mainly on lexico-grammatical artifacts based on recurrent patterns among various novels. By introducing the notion “motif,” which refers to the recurrent patterns, authors of Phraseology and Style in Subgenres of the Novel attempt to analyze styles of English and French novels through certain fixed patterns. The contributors offer insight into both genre-specific patterns and patterns found in novels in general. The recurrent patterns are relatively stable, so as the title indicates, the motif can be regarded as phraseology. Based on identifying certain frequently seen motifs, the phraseology in subgenres of the novel are explored in hope of distinguishing specific genres as well as analyzing linguistic and stylistic features. The researchers show that certain various recurrent patterns may serve as distinguishing features among various subgenres of the novel, and these specific motifs may enhance the literariness of related subgenres. Moreover, the authors identify motifs that can serve as genre markers, and through certain brief phraseologies, the subgenres can be distinguished accordingly and their stylistic features can be analyzed.The book is an edited collection, with each chapter written by different authors. The ten chapters that are well structured, starting with the general introduction to the approach and issues concerning general novels and then moving on to a comparison between specific genres. The detailed heuristic process and research findings will be illustrated below.Chapter 1 written by Iva Novakova and Dirk Siepmann, provides a general introduction to the PhraseoRom project, the English and French literary corpora compiled in the project, and to “motifs” as the kernel concept of the book under review. To begin with, former linguistic approaches to literature, phraseology, and idiomaticity are evaluated and the corpora methodology of the research is illuminated. The notion “motif” is introduced to bridge the gap between linguistics and literary studies. Moreover, the genre of fiction is divided into six subgenres for further study, namely crime, science fiction, fantasy, romance, historical fiction, and general fiction (8). According to specificity, frequency, and dispersion, recurrent lexico-syntactic trees (RLTs) are automatically extracted to detect and retrieve motifs. RLTs are lexical units linked by syntactic dependency relationships. They are constructed from statistically significant series of co-occurrences based on a statistical association measure, which in this context means calculating the log-likelihood ratio (LLR) using the keywords method. The Lexicoscope program is the automatic extraction tool used (Kraif). RLTs’ extraction starts with a lexical pivot (verbal or nominal) selected by frequency and dispersion criteria from among the authors composing the corpus and ends with producing a set of recurrent subtrees. To cite an example, “Pronoun/Noun He + V Lighted + DET a + N cigarette + PP (propositional phrase)” can be seen as a typical recurrent motif concerning a cigarette script. In Chapter 2, Dominique Legallois and Stefan Koch further illuminate the notion of “motifs” in detail, and illustrate various definitions among different disciplines. Motifs are called “recurrent element” in folkloristics, “stereotyped entity” in narratology, and “quantitative property representer” in automatic text processing, and those definitions can be roughly regarded as synonyms of “recurrent patterns.” Recurrent pattern is the key of motifs in this research. By analyzing motifs, the patterns specific to a textual genre can be defined, and the functions can be described.The method is applied to analyze certain motifs in various subgenres of English and French novels, in an attempt to identify recurrent patterns that may better distinguish the specificities of different literary subgenres. To make it plainer, motifs are regarded as genre markers.Chapter 3 analyzes key verbs and adverbial motifs among general English novels as well as their equivalents in French novels. The major concern is about manner adverb and motifs of motion and light. Based upon a large corpora and keyword approach, the researchers, Iva Novakova, Dirk Siepmann, and Marion Gymnich, find that adverbs are more frequently used in English than in French and that simpler binary collocation is preferred in English, whereas manner verbs are more complex in French and more periphrastic expressions are detected. The linguistic phenomenon empirically proves higher creativity in French literary language.In Chapter 4, the authors, Sascha Diwersy, Laetitia Gonon, Vannina Goossens, Marion Gymnich, and Agnes Tutin, compare the use of speech verbs in French and English novels. The speech verbs are drawn from the speech tag construction in three subgenres: fantasy, crime, and romance, and classified according to Speech Act Theory. The data reveal that English and French novels differ considerably in the distributional configurations of speech verbs belonging to different semantic categories. English has a universal usage of “to say” as the speech verb and tends to minimize the narrator’s intervention. What is more, the specificity of fantasy is highlighted in which the two languages do not share the same types of verbs. In French fantasy, expressive verbs are dominant, which contribute to an overemphasis on the character’s emotions and attitudes. The dominance of expressive verbs can also be regarded as the genre mark of French fantasy.Chapter 5 explores how depicting alcohol and tobacco consumption contributes to the portrayal of characters and the rendering of scripts in English and French novels since the 1950s. By retrieving RLTs related to alcohol and tobacco consumption, the authors, Fransis Grossmann, Marion Gymnich, and Dirk Siepmann, confirm that references to alcohol and tobacco consumption are used to state characters’ emotion and attitude or to announce or suspend the conversation. The cigarette and alcohol scripts seem to be quite universal among English and French, but French authors prefer the generic term “alcohol,” whereas English authors tend to use specific references, and the motif “take a sip of wine” is the genre marker for romance novels in English.Chapters 6 and 7 shift the concern to the way “fiction words” and collocations distinguish literary subgenres. Neologisms are features of science fiction, which refer to realities unfamiliar to readers. Fiction words are proposed to cover these neologisms, which have no referents in reality.Chapter 6 compares the usage of fiction words among English and French works of science fiction. The authors, Laetitia Gonon and Olivier Kraif, examine the words morphologically and semantically; the semantic field is applied for the classification of linguistic data and the syntactic distribution is analyzed for comparison. By using Lexicoscope and retrieving RLTs automatically, words that are annotated as fiction words are extracted. Semantically distributing, it is found that both in English and French fictions, the fiction words that most frequently appear are semantically derived from technological words. For example, bublepak, ornithopter, and passtouch in English; agrave, plastacier, and holorama in French. As for word formation, in French fictions, the types of words that most frequently appear are hybrid classical compounds like chromobéton and neuroséquenceur, purely classical compounds like sexomorphe and transvers, lexical creations like gleis and sukmen, and portmanteau words like plasticuir (plastique+cuir) and vocodeur (codeur+vocal); in English fictions, the types of words that most frequently appear are mainly popular compounds like flycar and passtouch, hybrid classical compounds like posthuman and nervoplex, and portmanteau words like fibdrop (optical+fiber) and graviscan (gravitational+scanner). Certain fiction words are applied universally, and two literary traditions may share a common background of fictional references.In Chapter 7, the authors, Vannina Goossens, Clemence Jacquot, and Susanne Dyka, differentiate science fiction and fantasy through semantic categorization of recurrent lexico-syntactic trees (RLTs). A semantic grid is introduced for retrieving linguistic data, more specifically, a semantic grid is settled to prerequisite that the RLTs have to contain a verb. To cite an example, “Communication: verbal” receives patterns that imply using the articulatory organs for communicating. The recurrent lexico-syntactic patterns of novels are classified in an exhaustive semantic grid, and the generic markers and certain stereotyped actions are thus highlighted. The semantic analyses make it possible to highlight the generic markers (the medievalism of fantasy, the salience of specific spaces, and certain so-called stereotypical actions). Moreover, categories of cognition define science fiction, whereas perceptual dimension defines fantasy.Chapters 8 and 9 focus on specific motifs defined as characterizing and structuring elements in the fictional world. Chapter 8 makes a comparison between general fictions and crime novels. By using the Lexicoscope to contrast the general fiction and crime novel corpora and extracting RLTs, the authors, Julie Sorba, Laetitia Gonon, Susanne Dyka, and Vannina Goossens, find that characters in general fiction tend to read more letters and books than those in crime novels. General fiction has a specificity in discursive functions, which deal with the relationships between the characters, their intimate life, and what they say about it; reading and writing are perceived as more static in French general fictions while more dynamic in English general fictions. Moreover, the identification of motifs and their discursive functions is believed to contribute to “an operating theory of genres” (p227).In Chapter 9, the authors, Susanne Dyka, Ludwig Fesenmeier, and Marion Gymnich, focus on the motif as the collocation “in a state of NP” in English novels and its equivalent in French novels. The data indicate that the usage of the collocation is more frequent in all the subgenres of English novels except science fiction, because the various subgenres prefer various states, respectively. Another interesting finding is that the motifs in general fiction are used with more verbs, adjectives, and nouns than in any other genres, which may indicate that in general fiction they are endowed with less-specific features. The motif here contributes a certain variety to the presentation of a character’s psychology. In Chapter 10, Iva Novakova and Dirk Siepmann give an overview of the book and emphasizes the interdisciplinary approach to subgenres studies.The book is cutting edge in its methodology by adopting advanced methods and tools, such as vocabulary growth curve (VGC) and RLTs. In particular, RLTs are adequate in connecting phraseology with the stylistic features in subgenres of a novel; the authors of the book attempt to base on the lexico-syntactic structure to retrieve certain motifs and relate to the stylistic analyses. Previous studies are generally concerned more about the comparison between genres of literary works, assuming that the style of an individual genre is unified, whereas the authors of Phraseology and Style in Subgenres of the Novel are more concerned with differentiating contemporary fiction subgenres between two languages, which may provide inspiration to the study of comparative literature. By setting certain fixed and frequently seen collocations and utilizing a corpus, the research has proven to be fruitful. The relatedness between the microlevel of recurring multiword units and the macrolevel of fictional scripts is established, and the motifs can be scientifically regarded as a certain kind of genre marker. Also, this study will be valuable and inspiring to automatic text classification in natural language processing (NLP).However, the studies of subgenres are not quite balanced, and the historical novel is not mentioned in the book; the motifs arrangements are not coherent among chapters, and it seems that they are selected for the sake of convenience. Moreover, some of the bibliographic information in the references are written and published in French and without English equivalents, so they are unintelligible to French-illiterate readers.