A and an in English Plays, 1580–1639

I

Discussing ’Tis Pity She’s a Whore in A Companion to Renaissance Drama, Mario DiGangi quotes the Cardinal’s “disturbingly glib and unfeeling assessment” of the action of the play, which is at once the play’s last line and its title: “’Tis pity she’s a whore” (578). The determiner a is an important part of the framing of the remark, which turns all Arabella’s individual and passionately felt experience into a dry, dismissive epithet. Later in his chapter DiGangi quotes the opening sentence of Perkin Warbeck, spoken by King Henry VII, which closes with his complaint that he is being treated “[a]s if we were a mockery king in state” (580). The royal “we” implies that he concentrates in himself absolute plural power, but the phrase “a mockery king” undoes all that: he is not being treated as the king, but as a king, not the single, concrete instance but an abstract member of a possibly well-populated category.¹

The determiner a, and its alternative form an, are inconspicuous elements of language—DiGangi does not mention the presence of a in his two quotations—but they are powerful markers of perspective, as the two examples show. A and an are very common in John Ford’s dialogue. In this his pattern fits a well marked trend. Playwrights use steadily more as and ans over the six decades from the 1580s to the 1630s, and Ford’s plays, written in the 1620s and 1630s, use on average more than any other playwright’s from the wider period.² Shakespeare’s frequencies are somewhere in the middle. His plays do not show any increase in use from early to late. His characters, though, are distinguished by different levels of use in ways that correspond closely with Robert Weimann’s distinctions between locus and platea modes of performance. Jonson, Middleton, and Webster also form part of a wider pattern of increasing a and an use, and this broader trend complements Weimann’s ideas in suggestive ways. A and an use proves to be remarkably useful in generalizing about progressive changes in early modern English drama.

II

The calculations are based on 190 plays, machine-readable versions of early printed or manuscript texts.³ Each play is assigned a year of composition, so the plays can be apportioned to decades.⁴ The median count of a and an increases in each decade from the 1580s to the 1630s: slowly at first, then quite rapidly, and then slowly again (Figure 1).

Figure 1.

Percentages of a and an in plays over six decades.

The median is the bold line in the chart. It represents the a and an count for the play in the middle of the distribution, with an equal number of plays higher and lower. The curve has a flattened “s” shape—each successive decade is higher than the last, but the rate of change varies, with a gradual increase at first, then a quite rapid advance, and finally a slower change. (The “s” curve has often been observed in language change, though the underlying explanation for this common pattern is by no means clear [Denison].) The other measures in Figure 1 are the minimum and maximum (the lowest and highest play count for a and an in each decade) and the first and third quartiles (the play with a count that puts exactly three quarters of the group higher than itself, and the play with a count that puts one quarter of the group higher). While the minima and maxima may simply reflect a single aberrant low or high value, the quartiles present a more stable indication of the counts of the group as a whole. The pattern for these four measures is less neat than the median—the maxima in particular fluctuate widely—but it remains true that values are much higher in the sixth decade compared to the first. Generally the range—the distance between minimum and maximum—becomes smaller over time. The change in frequencies over the period...