Is Data or Aren't They?

I haven't posted in a while on account of my midterms (which continue through next week) among other things, so I thought I'd take this opportunity to briefly comment on something which consistently irks me, namely, the mistaken notion that the collective noun "data" is grammatically plural. This is a widely spread notion and provides for that ugly but nonetheless far too common collocation "the data are," as, for example, on this page from the Wake Forest University Department of Biology. The roots of this misconception lie in the Latin analogue of our English word, which is a neuter plural past participle of the word for "give." In Latin, "data" are "givens," and thus, until actually quite recently, "data" was employed as a plural and meant "axioms." The OED cites the earliest usage of this sense from 1646. This, however, as any speaker of English will tell you, is not the most common meaning of the word "data."

The common modern meaning, which the OED sets as "Facts, esp. numerical facts, collected together for reference or information," shows up first in 1899. As it first came into usage, it may indeed have remained a plural in this sense; after all, people still learned Latin in those days. But after a century in the service of illiberally educated scientists, I can say with absolute certainty that in genuine usage the word is grammatically singular and plural only under the duress of pretension. I find no authority in the argument that the Latin word is plural: Latinitas non Anglice dicit, "Good Latin doesn't speak English." We have borrowed, for example, the Spanish word "siesta," which is grammatically feminine in Spanish, into our language. Should we then be saying "I approve of the siesta, it is a fine institution," or "I approve of the siesta, she is a fine institution." Obviously the latter is a little awkward, since only the biologically female are grammatically feminine in contemporary English.

Let us assume, however, that number is a different case from gender, since gender really only affects the personal pronouns, and take no note of the fact that number only affects present indicative verbs, and the preterite of "be." If we apply this steadfast principle of faithful Latinity to the whole language the results really become quite strange. Compare, for example, the verbs "translate" and "transfer." In Latin they are derived from different 'principle parts,' as they are called, of the same verb, the former being the perfect participle and the latter the present stem; the third 'principle part,' the perfect finite verb, transtuli, has not found its way into English. But if it had, what sentences we could make! "After that, Mr. Jones transtuled the works of Horace into English, having translated Ovid already. 'It is a goal of mine,' he said, 'to transfer all the great Latin poets into English.'"

The answer to this more absurd proposition, of course, is that different senses of a single Latin verb came into English by way of different stems. Now that they have become English, it is beyond silly to try to re-Latinize them. I fail to see how the case of "data" is much different. The sense of the word is undoubtedly plural, but in a collective way, which is hardly rare in English: navy, furniture, etc. The fact that the easiest synonym for "data" is "information" goes a long way, I think, towards showing the common sense of implementing it as a grammatical singular. English has simply used the word differently than Latin.

I will pass over the fact that neuter plurals always take singular verbs in Greek.