There exist speakers of English who have never been exposed to the classical languages and are convinced that every plural noun is derived from the singular by appending either `s', or `es'. It must have been in the interest of this group that Logan Pearsall Smith issued the following curious declaration:
No common noun is genuinely assimilated into our language and made available for the use of the whole community until it has an English plural, and thousands of indispensable words have been thus incorporated.On the contrary, there are many Greek and Latin plurals, e.g., data, agenda, media, opera, and crises, that have been absorbed unscathed into common English usage. On the other hand, one would only invite ridicule by insisting today on the abandonment of ideas, museums, specimens, and bandits in favor of ideae (or ideæ), musea, specimina, and banditti. In the middle ground are those nouns, such as memoranda and concerti, that appear to be awaiting final judgement. This includes a number of words taken from Hebrew, the structure of which is even less familiar to the general English-speaking world than that of Greek or Latin, Thus, bnai mitzvah and seraphim continue to struggle against bar mitzvahs and seraphs. Finally, we find various classical forms that remain in scientific usage—automata, formulae, foci, indices, etc.—but have been Anglicized in other contexts.
A matter of graver concern is the errant reconstruction of singulars from familiar classical plurals. Failure to identify the pluralization of the Greek suffix -is, for example, has led to such bizarre back formations as parenthese and indice. (I am unsure of the spellings, as I have encountered these barbarisms in conversation only, but both are pronounced with a final long `e'.)
When an irregular plural (a) is not readily recognized as such by uneducated speakers, and (b) happens to be the more common form of a noun, it is apt to be misused in place of the singular in the relatively rare contexts in which the latter is called for. Thus, as graffiti are customarily observed in clusters, a careless speaker is liable to refer to a single graffito as a graffiti. Even more absurd is the attempt to pluralize such a presumed singular, as in the King James Bible, where we find two references to seraphims.*
Naturally, numerous instances of this practice may be observed in the computer industry. (For further discussion of the abuse of language in this context, see MULTIPLE.) For instance, when a "bug" is discovered in a hardware design that has already gone to production, one hears of the need for an errata in the documentation.
A number of Latin plurals have become firmly established, through relentless abuse, as English singulars. Thus, agenda has evolved from the plural of agendum, which means a thing to be done, to a singular meaning a list of things to be done. But long after abandoning all hope of ever again seeing data used properly as a plural, I noticed that the linguistic fossil datum is making a curious comeback. Thus, a unit of information transferred between a processor and external memory is often called a datum. Alas, it is also not uncommon to hear of several such objects referred to as datums! This development warrants continued observation.