Book Reviews  159

were innovative in a number of ways, and Butler persuasively argues that
these innovations  —  the relative structural looseness of the masques, the
immediate printing of the masque texts, and the reinvention of masques as
‘triumphs’ — served a new political agenda. They helped to define and polarize political factions in the later reign as Charles’s glorious and triumphant
masques became more obviously distanced from his increasingly rigid and
inflexible political identity.
Admittedly outside Butler’s scope in this book are masques and entertainments performed away from the court space. The gap is a discernable one,
since Butler concludes the book with a suggestive discussion of the extent to
which Milton’s Masque Presented at Ludlow Castle (1634) ‘was constituted in
telling ways against the Whitehall masques’ (353). This conclusion hints at
an entire dimension of masques and entertainments occurring outside the
centre, and a consideration of this element in relation to the court masques
could certainly have enriched the study. Butler addresses this limitation,
noting that James Knowles’s forthcoming book will explore festival culture
occurring in the provinces and great houses. Appended to the book is an
exceptionally useful and detailed annals of masques and entertainments that
includes great-house theatricals as well as court entertainments. Butler’s book
is a valuable and elegant contribution to the study of the Stuart court masque
and will be integral to the reassessment of its complexly political nature.
Kirsten A. Inglis

Hugh Craig and Arthur F. Kinney (eds). Shakespeare, Computers, and
the Mystery of Authorship. Cambridge: Cambridge UP, 2009. Pp 234.
The title of Hugh Craig and Arthur F. Kinney’s new collection deliberately
juxtaposes two concepts often still considered incompatible. Can ‘computers’, with their binary and uncritical approach, really provide insight into ‘the
mystery of authorship’? Craig and Kinney’s neurologically-informed introduction suggests that creativity depends on a necessarily limited series of
patterns within the seemingly infinite human brain. Linguistic computing
depends on the belief that some of these patterns and habits of mind can be
identified and extrapolated using machines, allowing us to uncover a dis-

 160  Book Reviews

tinctive ‘authorial signature’ (12) even within relatively small sections of text.
Despite growing critical interest in the early modern theatre as a fundamentally collaborative and discursive environment, this volume is concerned with
‘disintegration’, the breaking up of collaborative texts to ascertain exactly
which words were written by which author, with the implied ultimate aim of
identifying the specific words that originated in Shakespeare’s mind. As plays
such as Double Falsehood, Thomas More, and Edward III progress to canonical status in the early twenty-first century, anxiety over the exact nature of
‘Shakespeare’ persists, and this volume offers a timely investigation into the
most contested areas of Shakespearean authorship.
Craig and Kinney’s methodology is deceptively simple, based on word
choice rather than collocations or literary style. ‘While an author can
always extend his or her range of active words — almost every new work
introduces some new words  —  the strong tendency always is to revert
to familiar and customary word usage’ (198–9), they suggest. The processes they utilize are thus as objective as possible, involving the literal
counting of occurrences of selected words while allowing for spelling and
usage variants. Where many other studies rely on a single methodology
Craig and Kinney’s team utilize two independent test strands in order to
verify conclusions. The first series of tests examines lexical words. The
team takes the two bodies of work for comparison, for example comparing Marlowe’s plays against Shakespeare’s plays, and generates lists of the
500 words most characteristic of each set relative to the other. This list
of 1000 words is then applied to the disputed text and its likeness to the
two test sets evaluated. As the authors point out (66), the strength of this
test is that it reduces the dangers of identifying only imitation or commonplace words by selecting those words that are not most frequent in
total but relatively most frequent and thus more unique to the test sample.
The second series of tests looks at function words, theoretically a relatively
subconscious element in writing. Again, the team identifies the function
words most specific to individual authors before considering their frequency in disputed texts. They take great care throughout to standardize
sample sizes and ensure that texts are compared on an equal basis, and
throughout emphasize that the results merely show greater or lesser levels
of probability: there is, and will always be, room for doubt.
Critics sceptical of the effectiveness of such methods are answered by rigorous control-testing at all stages against known samples of text. Only once the
team has established a strong identifying set of criteria do they apply the tests

 Book Reviews  161

to disputed texts. One of the collection’s most important and incontestable
achievements is its demonstration of consistency of word choice in the works
of specific authors, especially Shakespeare. A summary of the results will
be of primary interest to scholars. Shakespeare is confirmed as the author
of Hand D of Thomas More and the Countess scenes of Edward III, and
the reviser of his own King Lear in the Folio version. More controversially,
the team also grants Shakespeare the middle section of Arden of Faversham
and the 1602 additions to The Spanish Tragedy usually attributed to Jonson.
Marlowe is Shakespeare’s key collaborator on 1 and 2 Henry VI, writing the
scenes involving Joan of Arc and Jack Cade; and the volume further asserts
that the remainder of Edward III and Edmond Ironside bear little resemblance
to the works of any known dramatist, in the case of the former convincingly
rebutting recent arguments in favor of Marlowe and Kyd.
The results, presented in easily interpretable graphs, are persuasive on
paper but are not beyond question. Craig and Kinney pride themselves on
transparency; yet they offer surprisingly little of their intermediate data for
the benefit of critics hoping to reproduce their results. Why, for example,
do the authors supply no appendix detailing the lists of characteristic lexical
words used to differentiate authors? Without these lists, the reader is obliged
to trust that the lists generated for each test do indeed represent the most
characteristic words of each author without being affected by local anomalies
such as subject matter. A link to an online resource with full tabulations,
raw data and search criteria would be the next logical step in affirming these
results. The choice of plays that make up their test corpus also raises questions: why is The Taming of the Shrew, for example, dismissed from tests as
being of ‘disputed authorship’ (218)? Why are only 112 of the ‘174 surviving
well-attributed single-author plays’ (xvii) from the period used for comparison? These questions do not invalidate the volume’s results, but the inquiring scholar will wish to exhaust all possible variations in the test parameters
before accepting its conclusions.
Several texts also require more rigorous testing. In the case of The Spanish Tragedy the authors only present evidence that the lexical word test
favors Shakespeare over Jonson, Webster, or Dekker individually. Craig
claims that function word tests support this conclusion, but the results are
not shown. A lexical word test measuring Shakespeare against the collected
corpus of all other dramatists for the time places the additions in nonShakespeare territory. This result is passed over as an aberration; yet this
‘all-comers’ test is the bedrock of other chapters such as those on Henry VI,

 162  Book Reviews

Edward III and Thomas More. The results they present are thus suggestive
but severely limited, and again the reader wishes for an exhaustive series
of tests comparing every possible combination of dramatists in order to
demonstrate that the data continues to support the authors’ conclusions
as widely as possible. The authors thoroughly examine the external evidence pointing towards Jonson as the writer of the additions to The Spanish Tragedy to show an element of doubt in the attribution, but make no
attempt to account for how or why Shakespeare might be writing additions for a Henslowe play at the Rose in the late 1590s/early 1600s. Their
decision simply to present linguistic data for other scholars to interpret is
commendable, but only serves to reinforce the divide between literary criticism and linguistic computing that too often leads the former to ignore the
latter’s work. Hugh Craig’s chapter on Henry VI is more successful in this
regard, supporting his evidence for Marlowe’s hand with a consideration
of Joan and Cade as Marlovian, rather than Shakespearean, villains: their
low birth, their high ambition and ‘native wit and daring’ (67) relate them
especially to Tamburlaine.
The volume grants very little space to this kind of literary criticism, with
chapters given over instead to attribution history and the team’s tests. The
attribution histories usefully contextualize and provide precedents for the
team’s results drawn from literary criteria, particularly the early twentieth-century trend for parallel passages. Timothy Irish Watt’s chapters on Edward III
and Thomas More, though, are surprisingly error-ridden, twice rechristening
C.F. Tucker Brooke ‘G.F.’ (119) and misrepresenting other scholars’ work. In
a volume so dependent on accuracy and attentiveness to detail such slips are
discouraging. Any work of this nature must be scrutinized thoroughly for
flaws and potential holes; these difficulties, however, should not detract from
a volume that intervenes importantly in Shakespearean authorship disputes.
Craig and Kinney establish a standard of statistical rigor that all scholars in
the field should aspire to. The use of dual, independent tests to verify results;
the objective presentation of data; and the unbiased investment in whether
or not Shakespeare wrote a particular text are all essential if computational
methods are to be accepted by the literary community. The results presented
here are compelling; if the authors are willing to make the rest of their data
public and if exhaustive tests can be run to ensure that the results remain
equally convincing in every comparison of possible authors, then this volume
may well be pivotal in forcing scholars to take seriously the claims of Arden

 Book Reviews  163

of Faversham to Shakespearean authorship as well as opening The Spanish
Tragedy and the Henry VI plays to renewed scrutiny.
Peter Kirwan

Christa Knellworth King. Faustus and the Promises of the New Science,
c. 1580–1730. Farnham, Surrey: Ashgate, 2008. Pp 216.
In this study Christa Knellworth King attempts the very intriguing project
of tracing the significance of the Faustus myth from the chapbook sources
for Christopher Marlowe’s Doctor Faustus to the debased versions of the
tale in the farces and harlequinades of the early eighteenth century. The
various texts covered will be of great interest to anyone interested in the
ideological transformation of the myth in question. However, the relation of
this narrative to the emergence of scientific thought and practice in western
culture — that is, to the ‘promises’ indicated in the second half of King’s
title — remains finally a rather vexed question, partly (perhaps) because the
relation between magic and science remains so vexing historically.
What is always clear to the reader is enunciated pointedly at the end of the
study: ‘the creative energy that radiates from the depiction of Faustus’ adventures suggests that all versions discussed here were on the side of Faustus,
even if the arguments in his favour are expressed in parallel with those of the
prophet of doom’ (186). From the chapbooks on, King argues persuasively
for a sympathetic reading of the hero’s aspirations; the ‘subversive’ meaning
consistently stands out even when the ‘ostensible purpose’ is ‘a cautionary
tale’ (184). In addition to this persistent demystification, King’s argument
heightens critical interest by showing how ‘the Faustus narrative … turns
into a study of human attempts to deal with responsibility and moral selfdetermination’ (185). I was particularly intrigued by King’s suggestion that
the comedy of the later, farcical versions of Faustus is not simply gratuitous
since such versions ‘celebrate the protean hero whose elusive appearance guarantees his success in an upwardly mobile society, which is to say he illustrates
the aspirations of his period’s increasingly powerful middle classes’ (168). To
increase a sense of critical continuity, this aspect of King’s argument might
be linked more closely to Marlowe’s own hero, born ‘base of stock’ but hugely