This bad news for the colon was excellent news for the semicolon, however. It was the semicolon that writers substituted for its unstylish progenitor, and the semicolon soon surged through sentences at such a pace that it gobbled up not just most of the colons but a large share of commas as well. The semicolon had become so fashionable by the 1840s that Goold Brown leveraged its appeal to implore writers to reconsider the now-neglected colon. ‘But who cannot perceive,’ begged Brown, ‘that without the colon, the semicolon becomes an absurdity? It can no longer be a semicolon, unless the half can remain when the whole is taken away!’ Brown pointed out that there was really nothing wrong with the colon anyway; after all, he argued, colons, too, were ‘once very fashionable’. (Brown himself was not, however, immune to the seductions of the semicolon: the first sentence of his massive grammar compendium contained seven of them, and nary a colon.)
Although Brown appealed to fashion in his plea for the colon, his argument offered a glimpse into the future of punctuation – the future we’ve inherited – in which logic reigns supreme in guides to punctuating. By the time renowned British grammarian H. W. Fowler published A Dictionary of Modern English Usage, in 1926, Brown’s halves and wholes and absurdities had been pressed by Fowler into sterile language reminiscent of a reductio proof in mathematics, a curtly logical formulation in which neither fashion nor taste nor passion nor prosody had any part:
The use of the semicolon to separate parallel expressions that would normally be separated by commas is not in itself illegitimate; but it must not be done when the expressions so separated form a group that is to be separated by nothing more than a comma, or not separated at all, from another part of the sentence; to do it is to make the less include the greater, which is absurd.
The semicolon had been transformed. Before the 1800s, it had been a pause. By the early 1800s, grammarians began to describe these pauses as means to delineate clauses properly, such that punctuation served syntax, with its prosodic and musical features secondary. By the mid-1800s, guided by a new generation of grammarians, grammar was tiptoeing towards a natural science model, deriving its rules from observation of English and teaching those rules to students through exercises in which they would be guided to make the same observations and draw general conclusions from them in the form of rules.
This process of moving from particular examples to general principles is known as ‘induction’. Faced with the shift towards an inductive method modelled on science, mid-century grammarians waffled on the proper place of punctuation in their guidebooks. Was punctuation part of orthography, the study of how a language should be written? Was it part of prosody, how language should sound? Or did punctuation fall under the heading of syntax, the study of how language should be structured? The problem sparked vigorous debate. If punctuation were to be part of prosody, how could it be taught with the properly scientific inductive method? A student could hardly be expected to inductively derive rules from the rich, subtle, and infinitely varied rhythms that punctuation created in texts!
Uncertain of punctuation’s proper classification, some grammarians simply left it out of their books entirely, sidestepping the problem of making rules that would be appropriate for an observation-based, natural science of language. Where punctuation was included in mid-1800s manuals, however, writers were given ample opportunity for semicolon use. In George Payn Quackenbos’s 1862 An English Grammar, four possibilities for semicolon usage were given, and students were instructed to use a semicolon (not a colon, as we moderns would) to introduce a list of items.
Quackenbos’s rules for the semicolon.
Nowadays all style guides advise against using a semicolon to link together an independent and a dependent clause. In other words, the parts of a sentence separated by a semicolon must be able to stand on their own grammatically, so that they could be complete sentences if they wanted. But most nineteenth-century grammarians considered it perfectly acceptable to link together an independent and a dependent clause using a semicolon. If the author thought it sounded right, no need to meddle. As the grammarians’ push towards logic, natural science, and induction continued, however, punctuation became decoupled from prosody and personal preference. By the 1880s, grammarians distinguished between rhetorical pauses, which were akin to pauses in speaking and were not to be marked by punctuation; and grammatical pauses, which required punctuation in order to make the structural and logical attributes of a sentence clear.
There were a few holdouts who resisted the new model of grammar. W. C. Fowler, for instance, advocated applying punctuation marks to signal rhetorical pauses, and in his 1881 grammar it was still permissible to use a semicolon between an independent and a dependent clause. But by his own admission, Fowler was the exception rather than the rule: by the time Fowler published his book, grammarians were treating the semicolon like a controlled substance. They generally prescribed them only for use between independent clauses, or to separate items in a list that were long enough to be subdivided with commas. An 1888 report on punctuation by the California State Board of Education called attention to the fact that there was ‘but one use of the semicolon’ in its lessons, its function restricted to separating independent clauses that contained commas. One hundred and thirty years had passed between Robert Lowth’s grammar, which had envisioned punctuation marks as musical elements in prose, and this report, with its strict logic-based rules. In that time, the function of the semicolon had grown narrower and narrower until it had been whittled down to one or two applications.
This narrowing, as we’ve seen, happened in large part because grammar books were part of greater cultural and aesthetic trends in the Western world that privileged natural-scientific knowledge above other types of study; and in responding to those pressures, grammarians devised systems of punctuation that fitted comfortably with the ideals and the formats employed in science textbooks. Where had a century of rules, of grammars labelled ‘true’ and ‘real’ and ‘improved’, of grammars sidling up to natural science – where had all this got the generations of English speakers who had grown up immersed in books devoted to these principles? Into trouble, that’s where, according to commentators. ‘Where is the man that can tell why a comma is inserted instead of a semicolon; a semicolon, instead of a period [full stop]; a colon, instead of a period?’ asked an 1899 American newspaper editorial. ‘And yet, the science of punctuation is almost as definite as the science of mathematics, and, with few exceptions, a reason can be given for every mark inserted.’ The rules were definitive, and yet their application varied wildly, and this divergence was apt to haunt a person all his or her life. As the Boston Daily Globe sombrely intoned:
From childhood’s earliest hour, or from that early hour in which one graduates from pothooks and hangers, and his wavering hand begins to