The plural of virus? Latinate plurals reconsidered

Jan 19, 2012 by

[Thanks to Cynthia Typaldos for bringing this topic to my attention!

The cartoon to the left is from this site]

A recent article on by Robert Andrews states:


“Google (NSDQ: GOOG) already operates its own legal book sales elsewhere on Android Market and in Google Books. But its controls on approval in the main app market are minimal. Previously, it has reacted post-publication to remove apps that include malware, for example, exposing users to mobile virii.”

Virii?! This is intended to be the plural form of the Latinate noun virus. However, it is a perfect example of hypercorrected (i.e. “more Catholic than the Pope”) Latinate plurals. Let’s consider this topic more closely.

Latin has contributed many nouns to the English language, and many of them still form plurals in the Latin-derived fashion. For example, the plural of crisis is crises (not *crisises) and the plural of corpus is corpora (not *corpuses). Similarly, the plural of phenomenon is phenomena. Other Latinate noun have both Latin-derived plurals and more regular English counterparts: for instance, the plural of formula is both the Latin-derived formulae and the regular -s plural formulas. Similarly, the plural of forum is both fora and forum, and the plural of schema is both schemata and schemas. Interestingly, in all these instances of alternative plurals, the regular -s plurals are shown to be more common by Google searches. For example, a search for formulae brings up 3,180,000 hits, whereas formulas comes up with 15,000,000 hits.

But while Latin-derived plurals are appropriate for some Latin-derived nouns, in other cases Latin-style plurals are sometimes used even though the nouns would not have that plural form in Latin itself. Case in point is virii (or viri), the putative Latin-derived plural of virus. In fact, the word virus in Latin was a mass noun (not unlike English footwear or furniture): while referring to entity that can be conceived as consisting of individual units (e.g. furniture consists of individual pieces such as tables, chairs etc.), virus was conceptualized as a mass/substance. And as a mass noun, it did not have plural at all (note that there is no *furnitures in English).

Another example of a hypercorrected Latinate plural is the word syllabus. Numerous times I have heard my colleagues in the academic world report having prepared their course syllabi. But this form is incorrect too, as far as Latin itself is concerned: in Latin the plural of syllabus is syllabūs, with a long rather than short vowel in the suffix. Similarly, the plural of octopus is not *octopi, as this noun is a Latinized form of Greek ὀκτώ-πους ‘eight-foot’. The theoretically correct form octopodes is rarely used, so unless you want to sound pretentious, use the regular English plural octopuses.

This brings me to the much debated issue of data: is it singular or plural? In Latin, from which the noun derives, the form data is the plural of the singular datum. But the singular datum is not used in English, hence the confusion. So should it be the data is … or the data are? Usage differs depending on the register and the dialect. In popular usage the more frequently used form is the data is/shows/… thus treating data as the singular. In scientific usage, however, there is a difference between the British and the American usage: in the UK the preferred form is the data are/show/etc… (not unlike the British use of collective nouns, as in the police are…). In the US, you are more likely to encounter the data is/shows/… in scientific usage and popular usage alike.

In fact, Latin is not the only language that has contributed plural nouns that have been refashioned in English to be singular nouns. For example, Italian has contributed the following i-plurals all of which have become singular in English: biscotti, graffiti, paparazzi, panini and of course spaghetti. In Italian, the proper singular forms are biscotto, graffito, paparazzo, panino and spaghetto (of course, one rarely talks about one spaghetto in Italian, but one can).In English, however, the Italian plural forms serve as the singular, with some of the nouns acquiring the doubly-plural is-forms, as in paparazzis and paninis.

Related Posts

Subscribe For Updates

We would love to have you back on Languages Of The World in the future. If you would like to receive updates of our newest posts, feel free to do so using any of your favorite methods below:

  • Guest

    Nice. Some remarks though:
    Phenomenon and schema are actually Greek nouns, and so their plural forms. It seems English has borrowed the word “schema” more then once: another borrowing is “scheme”. The latter form lost the final vowel and has regular plural form just like other Greek nouns of this declination: theorem, program etc. (there is no hyper-corrective usage “theoremata, problemata”). It also had a similar bifurcation in Latin: the original Greek plural “schemata” was not the only form, there was also Latin new-formation “schemae”.

    Originally spaghetto is the diminuitive of “spago” – string, cord. So as with almost any type of pasta, the singular use is fine mainly in its original meaning: farfalla – butterfly, penna – feather etc.

    • Asya Pereltsvaig

      You are absolutely correct on phenomenon and schema being Greek. As far as I know they came into English via Latin. I should have said this more clearly in the post.

      And great point on the different types of pasta!

  • John Cowan

    In fact, syllabus is not a legitimate Latin word, nor does it belong to the fourth declension.  It first appears  in 1470 in a printed edition of Cicero’s letters to Atticus as a misreading of sittybas, a Latinization of the Greek word σιττύβας, the accusative singular of σιττύβα, meaning a label or title-slip on a book (origin unknown).  Neither the OED nor Lewis and Short (I don’t have access to the OLD) records any plural form other than syllabi.  The Greek word σύλλαβος to which older sources attribute the term does not exist.

    • Asya Pereltsvaig

      Thank you for sharing this information, John! There does seem to be some disagreement as to the origin (and thefore proper plural form) of syllabus — to be further investigated.

    • Rick

      The best article on this question that i have seen is here:

      I don’t know enough to judge the merits of the writer’s case, but I’m thinking it really ought to be referenced in the Wikipedia article on this question.

  • Twitley

    Excellent points.  But some nouns can be used in mass or count sense. e.g. ‘poisons’.  Is there a large enough surviving corpus of Latin to determine whether ‘virus’ might have been so used?

    • Asya Pereltsvaig

      It’s an excellent question but alas I don’t know the answer…

    • John Cowan

      The small number of references I was able to check at Perseus are all in the singular, as would be expected for a mass noun.

      By the way, statisticians do use datum, as in this little dialogue: “I am not just a statistic!”  “That is true; you are a datum.”  So do geographers in a different sense: “a set of reference points on the Earth’s surface against which position measurements are made and (often) an associated model of the shape of the Earth to define a geographic coordinate system.”  In this use, the plural is datums.

      • Asya Pereltsvaig

        Thanks, John, for your insightful comment, as always!

  • Alex

    Another example (particularly relevant in this political season) is caucus.  Some sources claim the origin is medieval Latin, itself from medieval Greek kaukos (some form of drinking cup), yet others (better supported as I recall)  from an Algonquinian word for counsel. Only the first would justify the latinate *cauci, caucuses seems a better plural.

    • Asya Pereltsvaig

      I was thinking about this word too, especially in light of our series of posts on the Caucasus on GeoCurrents:

      My etymological dictionary confirms the Algonquian origin of the “caucus”…

  • Olga Kagan

    I keep having trouble with formulae/formulas. Personally, I perceive formulae as the correct form (so I’ve learnt), but I see that more and more people use “formulas” in scientific articles. So I start feeling uncomfortable when saying “formulae”. I get an impression that when people see me use this form, they think I’m showing off.

    • Asya Pereltsvaig

      A google search indicates that “formulas” is used far more frequently than “formulae”, so the trend towards regularization is clear. I wouldn’t say that your use of “formulae” means that you are showing off, but you are certainly using the more old-fashioned (some now would say “old fashion”) form!

  • Mc Apostoli

    All dictionary entries I’ve looked up have had ‘viri’ as the plural of virus. I don’t think this is a mass noun. 

    • Asya Pereltsvaig

      It is not in English but was in Latin.

  • Martingarrod

    data is clearly a mass noun. one item of data is an anecdote.

    • Asya Pereltsvaig

      Yes, “data” in English is a mass noun (in Latin it wasn’t), but that doesn’t solve the problem of number as mass nouns can be either singular or plural in English: witness “the rice IS cooked”, but “the beans ARE cooked”.

  • N. Stanzione

    F.W.I.W., the fact of something being a mass noun in Latin does not preclude its having a plural (e.g., the more common synonym for virus, venenum, has a well-attested plural). However, since virus is neuter, the proper plural would be vira. Viri is a plural of vir (man, hero).