So-Called “Pronouns” in English

Pronouns include forms such as I, we, you, he, she, it, they, as well as their inflected forms such as me, him, her, them and reflexives, such as myself, yourself. It is traditional to think of pronouns as replacing phrases containing a noun (noun phrases).

For example, consider sentence (1) I ate the red apple. We can replace the red apple with the pronoun it, producing sentence (2), with broadly the same meaning: (2) I ate it. The pronoun it replaces the whole noun phrase the red apple, not just the noun apple.

A different proposal

In his 1966 paper On “so-called Pronouns in English”, Paul Postal argued that a pronoun is not a substitute for a noun phrase, but a special form of the definite article used before a silent noun. So he analyses sentence (2) as looking like (3) I ate it-(thing). In (3), thing is a noun that remains unpronounced and it is the form of the definite article that is used before unpronounced third person inanimate nouns.

Postal wrote itthing. For legibility, I have placed thing in strikethrough and in brackets to highlight that it is silent. I’ve also introduced a hyphen to separate the two forms it and thing.

Most of Postal’s examples use one as the silent noun, producing forms such as I-<one>, he-<one>, we-<ones>, them-<ones>. (Postal wrote them as ione, heone, weones, themones.) But sentence (3) contains it-(thing) because Postal notes that it sounds a little more natural than it-(one).

Advantages of Postal’s proposal

Postal identifies the following advantages of his proposal:

  • it explains why the normal definite article the cannot appear before the pronoun one.
  • it provides a convenient way to analyse reflexive pronouns.
  • it explains why pronouns can sometimes appear before nouns.
  • it is consistent with the fact that the silent noun is actually pronounced in some cases.
  • it helps explain some similarities between the form of the definitive article in English and some pronouns.

When is the one possible?

Postal notes that sentences (4a-c) are acceptable but sentence 4(d) is not acceptable:

(4a) I ate the apple Schwartz gave me.
(4b) I ate the one Schwartz gave me.
(4c) I ate the apple.
(4d) * I ate the one.
(the conventional asterisk * indicates an unacceptable sentence.)

Sentence (4b) shows that one can replace a noun phrase that introduces a restrictive relative clause. But sentence 4(d) shows that this replacement is impossible if no such clause follows.

Postal’s proposals provides the beginning of an elegant account of why sentence (4d) fails: he stipulates that one is silent if no restrictive relative clause follows— and when that occurs, the definite article is not the but it. But his explanation is not a full account. He doesn’t explain clearly why one is silent before a restrictive relative clause but not in other cases

Reflexive pronouns

Reflexive pronouns are forms such as myself, yourself, herself, itself, themselves. Postal’s proposal analyses them in a way that neatly parallels his treatment of other pronouns. In English, reflexive pronouns are a transparent combination of 2 components (2 morphemes): for example, itself is it-self. This is just like the phrases discussed above, except that the noun self is pronounced in this case. Indeed, self sometimes behaves as a completely separate noun, like all other nouns.

Postal doesn’t discuss reflexive pronouns in other languages. These are often not made up transparently of two separate morphemes, as they are in English.

Pronouns before nouns

Postal notes that pronouns sometimes occur before normal nouns that are pronounced, for example in phrases such as we women, you guys. Such phrases require additional explanation if pronouns are analysed as replacing noun phases. They require no further explanation under Postal’s approach, which treats the pronouns simply as definite articles preceding the (pronounced) nouns in the normal way.

His proposal does, though, need more discussion of why these nouns are pronounced.

Pronouncing the silent noun

Postal notes that the hypothetical pronoun stem one does actually show up sometimes, for example before an adjective in phrases such as we great ones. It is not surprising that a definite article (in this case we, on his analysis) can precede a phrase comprising an adjective followed by a noun.

Similarities in form

Postal notes that his proposal could help explain some similarities between the form of the definitive article in English and some pronouns. Very few English words start with the voiced inter-dental fricative consonant at the beginning of the word <the>, for which the symbol in the International Phonetic Alphabet is /ð/. Among them are the pronouns they, them, their and theirs. If those forms are analysed as a form of the definitive article used in particular environments, it will be possible to describe more concisely the characteristics of the words that can start with /ð/.

Postal suggests that this argument is even stronger in some languages where pronouns look even more like definite articles. He cites German (er / der; and sie / die) and Spanish (el / el and ella / la).

Kayne, Leu and Zanuttini (2014) comment that Postal’s proposal makes the identity of pronouns and definite determiners found in Romance languages (eg Italian la, lo, Spanish el) seem entirely natural.


Postal does concede that his proposed ‘special rule’ deleting the pronoun stem (such as one) ‘seems a bit ad hoc’. He doesn’t work through all the mechanics of how it works or provide any real explanation for it. But it does permit an elegant analysis of:

  • a pronoun—as a definite article followed by a silent noun.
  • a pronoun introducing a relative clause—as a definite article followed by one.
  • a reflexive pronoun—as a definite article followed by self.


My summary above is based on an excerpt of Postal’s paper, reprinted with an introduction and some thought-provoking questions, in An Annotated Syntax Reader: Lasting Insights and Questions, edited by Richard Kayne, Raffaella Zanuttini, and Thomas Leu (2014)

Leave a comment

Your email address will not be published. Required fields are marked *