Arnold Zwicky's Blog

Over on ADS-L a few days ago, Charlie Doyle reported on a usage that struck him as odd:

I have just read a student essay in which about 10% of the sentences begin with “As well [comma]” — instead of “Moreover” or “Furthermore” or “What is more” or “Also.” Our students (evidently) have been taught to stick in lots of “transitions,” but sentence-initial “As well” strikes me as abnormal. As well, the essay didn’t have much content.

(Vexingly, I was sure I’d been through this discussion before, but I can’t find a trace of it.)

In any case, about 20 years ago, a friend of mine who edits material (including scholarly writing) for publication complained to me about sentence-initial as well (which she was finding surprisingly often in otherwise very polished writing — surprisingly, because she associated the usage with vernacular speech and unpracticed writing), and I recalled similar complaints from composition teachers at Ohio State in the 1970s.

[I’m referring to the usage in question as “sentence-initial as well” for short, but the label can’t be taken as a description of the phenomenon. I’m looking only at as well as a complete constituent (not in as well as X, even when this is sentence-initial), set off prosodically in speech from the clause that follows (and set off with a comma in writing) and serving as a sentence adverbial syntactically and as a discourse connective pragmatically.]

Asking around a few years later, I gathered similar complaints, but also baffled responses from Canadian writers, who found the usage entirely natural. Bryan Garner, as it turns out, was already on the case.  from Garner’s Modern American Usage (3rd ed.):

as well. When used at the beginning of a sentence, this phrase has traditionally been considered poor usage. [It’s not in MWDEU, so I don’t have easily available cites from the advice literature about the tradition in question.] But in Canada it’s standard as an equivalent of Also, … or In addition, …

Garner gives three examples from Canadian publications — Toronto Star, Vancouver Sun, MacLean’s — edited for AmE, with And in two cases and Also, in one.

The qualms that some AmE speakers have about sentence-initial as well don’t carry over to sentence-final occurrences, as in examples like:

There were high winds last night. It rained(,) as well

The case is interesting, because it’s hard to see how you could object to sentence-initial as well on logical grounds (while accepting in addition and also and the formal idioms moreover and furthermore in this position, and also accepting sentence-final as well); the objection clearly arises from some perception about the kinds of people who use it and to the contexts in which they use it — as for sentence-initial discourse connective plus (which some handbooks simply deprecate, while others mark it as too informal for formal writing).

(Possibly Doyle’s student had somehow concluded that as well was more elegant than also or in addition. Or maybe the student found it natural in speech and carried it over into writing.)

There are similar judgments about sentence-initial discourse marker too (but not the sentence-final variant). Garner declares:

It is poor usage to begin a sentence with too (= also), although there is a tendency in facile journalism to use the word this way.

He advises also instead, adding that:

Words such as moreover, further, and furthermore are also serviceable in this position.

MWDEU takes a more measured approach, saying that other initial connective adverbs are more common “but too is not incorrect.”  It notes that the OED in 1913 claimed that too is “rarely, now never, used at the beginning of a clause”, but the later Supplement indicates that the usage has been revived in the 20th century.  MWDEU concludes that “No grammatical objection can be made to it”, but presumably because of its relative rarity it sounds peculiar to some people, so some handbooks discourage it.

Here’s an example of sentence-initial too in formal writing:

Modern nation-states have depended on the systematic collection of demographic data to manage public health, assess economic progress, and craft social policies.  In the United States, the official census initiated in the 1780s was coincident with the nation itself.  Too, already by the turn of the eighteenth century, a variety of nonstate enterprises were tabulating birth and death rates, or “vital records,” in order to track epidemics and devise insurance tables. (Sarah E. Igo, The Averaged American: Surveys, Citzens, and the Making of of a Mass Public (2007), pp. 6-7)

This entry was posted on March 25, 2011 at 9:59 am and is filed under Discourse, Modification, Usage advice. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.