著者

太田 一昭

出版者

九州大学大学院言語文化研究院

雑誌

言語文化論究 (ISSN:13410032)

巻号頁・発行日

no.40, pp.1-14, 2018

---

Hope's The Authorship of Shakespeare's Plays, published over two decades ago, is one of the most highly cited books in recent Shakespeare attribution studies. The present paper reviews the book, arguing that although it offers a valuable documentation of how playwrights responded to the rapid rate of change in English grammar in Renaissance England, Hope's method to combine linguistic and statistical analysis is only of limited use for authorship attribution. The Authorship of Shakespeare's Plays explores the authorship of Shakespeare's collaborations with Fletcher and Middleton, and the "apocryphal" plays. Hope attempts to determine the authorship of these plays based on what he calls "socio-historical linguistic evidence." He carries out a comparative investigation of the differences between playwrights in the usage of the auxiliary do, the relative pronouns who, which, and that, and the second person pronouns thou and you, and argues that these three forms of linguistic evidence can be used as authorship tools. Of these three, he decides that auxiliary do evidence serves as the most powerful tool that can be used for distinguishing playwrights. He examines the percentage of the "regulated" (now standard) use of the auxiliary do in each scene of his target texts, Shakespeare's "collaborative" plays, as well as in each of his control texts, Shakespeare sample 1 (six early plays of Shakespeare), Shakespeare sample 2 (ten late plays), and works by other playwrights (twenty-seven plays by five writers). He finds that the average regulation rate of the auxiliary do for the two Shakespeare samples (82%) is much lower than that for any non-Shakespeare sample. For example, the average regulation rates for the Fletcher sample and the Middleton sample are 92 percent and 90 percent respectively. Hope uses these collected statistical data to determine the authorship of individual scenes as well as of individual plays. Hope's approach to authorship attribution based on linguistic evidence has been highly regarded by leading Shakespearean scholars. Brian Vickers calls his method using auxiliary do evidence "a new and powerful authorship tool" (Shakespeare, Co-Author 121), and Tiffany Stern refers to his "tracing different early modern forms of auxiliary 'do'" as a "brilliant" analysis ("Fletcher and Theobald as Collaborative Writers" 130). The problem is that Hope's comparative analysis using the auxiliary do is capable of distinguishing playwrights in some cases but not in others. For example, the overall regulation rates for the Shakespeare samples (82%) and the Marlowe sample (87%) are different enough to discriminate the two writers. However, if we compare the regulation rate of Shakespeare's Richard II (83%) and that of Marlowe's Edward II (85%) — two English history plays analogous in subject matter — the difference is not great enough to be statistically significant. Hope's comparative method using do evidence can be even more problematic with individual scenes than with individual plays. He states that "a significant sample represents a scene of fifty or more tokens" and that "the degree of certainty of any ascription will increase with sample size" (Hope 24). However, his method does not work effectively with some of those "longer" scenes that have "fifty or more tokens." The scene by scene regulation rates for Shakespeare's "collaborative" plays as well as for the control texts are given in the Statistical Appendix (Hope 156-76), and the figures for many of the scenes of "fifty or more tokens" which critics claim Shakespeare may have written appear much closer to the figures for non- Shakespeare plays than to those for the Shakespeare samples. For example, three of the five scenes of The Two Noble Kinsmen which many critics believe were written by Shakespeare show more Fletcherian than Shakespearean regulation rates (1.01=90%, 5.01b=89%, 5.04=90%). The regulation rates for two of the three scenes of Edward III (1.02=91%, 2.02=90%) which many Shakespearean scholars agree were written by Shakespeare are significantly higher than the average rate for the Shakespeare samples (82%). This indicates that Hope's comparative analysis based on auxiliary do evidence is not always a dependable tool to identify Shakespeare's hands in his "collaborative" plays.