On Null subjects and related phenomena

Aug 25, 2010 by

In the latest posting I stated that Universal Grammar is a catalog of significant differences between languages rather than being a “one-size-fits-all” grammar. These significant differences are formulated as parameters, which are typically binary (i.e., allowing for two possible settings). We’ve examined the Headedness parameter (does the head of a phrase precede or follow its complement? left or right?) and the Verb-Second parameter (is the finite/tensed verb required to appear in the second position in the clause? yes or no?). Here let’s examine another parameter which has to do with the appearance of a subject rather than with the position of the verb.

One significant way in which languages differ is in whether the subject of a simple tensed sentence can be missing. For example, in Italian one can get away with Parlo italiano (literally, ‘speak Italian’ meaning ‘I speak Italian’); the first person singular nature of the understood — but unpronounced — subject is encoded through the form of the verb, specifically the affix -o. Obviously, in English we cannot say Speak English to mean ‘I (or he or we or they) speak English’. Not only do we need to say the subject explicitly when it refers to someone or something; we need some sort of pronounced (or “overt”) subject even when it does not refer to anything, as with It rains. In this sentence, it does not refer to any object or abstract notion. It’s not that “something” (the sky, the cloud, the rain) rains. In fact, in Italian no overt subject appears in such sentences: Piove. Indeed, it is impossible (or “ungrammatical”) to say in Italian the literal translation of the English It rains (*Ciò piove).

But the real strength of linguistic parameters is in that they don’t just describe one particular difference but account for several — superficially unrelated — phenomena. In the case of the Null subject parameter, it accounts not only for the obligatory or optional appearance of subjects in simple tensed sentences, but for a number of other phenomena involving appearance or non-appearance of subjects (in certain positions).

One such phenomenon is the optional appearance of subjects after the verb in simple sentences: in Italian it is allowed, as in Verrà Gianni (literally, ‘will-arrive John’), but in English it is not (hence, *Will arrive John is ungrammatical).

Moreover, not only placeholder (or “expletive”) subjects are required in English (but not in Italian) with the so-called weather-verbs (‘rain’, ‘snow’, etc.), they are also required in English — but cannot be used in Italian — when a subject that is itself a sentence appears sentence-finally. Let’s look at this more closely. Take the English sentence That Louisa will not leave is clear: here, the subject of the sentence as a whole is itself a mini-sentence (or “clause”) that Louisa will not leave. Another possibility is to put this subject clause at the end of the whole sentence: It is clear that Louisa will not leave. Note that here another “placeholder subject” it must appear in the canonical subject position, at the beginning of the sentence. It is impossible to say Is clear that Louisa will not leave. But once again in Italian things are different: Che Louisa non partirà è chiaro (lit. ‘That Louisa not will-leave is clear’) is a good sentence, as is È chiaro che Louisa non partirà (lit. ‘Is clear that Louisa not will-leave’). A placeholder subject ‘it’ is not only not required, as in English, but is not even possible: *Ciò è chiaro che Louisa non partirà (lit. ‘it is clear that Louisa not will-leave’) is ungrammatical. (Linguists mark ungrammatical sentences with an asterisk before the ungrammatical form).

And this is not all. In languages of the Italian-type subjects can be questioned even if there is a “complementizer” (a word like English ‘that’). (The original position of the subject is marked by __.) So in Italian one can say Chi credi che __ verrà? (lit. ‘Who you-believe that will-come?’); in English the corresponding *Who do you think that __ will come? is ungrammatical (one can say instead Who do you think will come?).

Other Null-subject-related phenomena include agreement of copula verbs and the possibility of restructuring, but we won’t go into that here.

To recap, English and Italian differ in a number of ways that involve the appearance of subjects. But it is not just English and Italian. Other languages pattern with either English or Italian — crucially, in all the abovementioned ways. Spanish is like Italian in allowing missing subjects (e.g., Baila bien lit. ‘Dances well.’); allowing post-verbal subjects (e.g., Llego Maria ayer a los doce lit. ‘Arrived Maria yesterday at noon.’); allowing questioning subjects immediately following ‘that’ (e.g., ¿Quién dijiste que ___ vino? lit. ‘Who did you say that came?’); and not requiring placeholder subjects when the subject-clause appears sentence-finally (e.g., Me parece que Juan tiene hambre lit. ‘To me seems that Juan has hunger.’).

French, on the other hand, is just like English in not allowing missing subjects (cf. the ungrammatical *Dance bien lit. ‘Dances well’); not allowing missing subjects even when they do not refer to anything (cf. the ungrammatical *Pleut lit. ‘Rains’); not allowing post-verbal subjects (cf. the ungrammatical *Arrivera Jean lit. ‘Will-arrive John’); not allowing missing subjects when the subject clause appears sentence-finally (cf. the ungrammatical *Me semble que Jean a faim lit. ‘To-me seems that John is hungry’); and not allowing questioning a subject immediately following ‘that’ (cf. the ungrammatical *Qui veux-tu que ___ viendra? lit. ‘Who want-you that __ will-come’).

Two things must be noted here: first, all languages pattern with either English or Italian. There is no “mix-and-match”; all these phenomena — although superficially distinct — are tied together. And it is not a priori clear why that should be so. Second, the situation with the Romance languages described above is particularly interesting because historically Spanish is closer to French than to Italian, yet Spanish patterns with Italian with respect to “null-subject phenomena”. Not only that, but historically, French used to be a well-behaved Romance language like Italian, Spanish, Romanian and Latin, allowing null-subjects (or missing subjects). What also becomes clear from the historical record is that French switched in all the relevant respects — at the same time!

Thus, the theory of parameters that bundles certain phenomena together can account for the bounds of variation (languages that are uniformly English-type or Italian-type are allowed but not “mix-and-match” languages), as well as for language change.


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: