Some English speakers allow a possessive relative that’s (sometimes written as thats) in place of whose in sentences like (1).
1) a. The girl that’s painting was bought is a popular artist.
b. The girl whose painting was bought is a popular artist.
In (1a), that's seems to be serving the same function as whose in (1b). While most American English speakers find whose to acceptable in such contexts, only some speakers use or accept that's in this use.
Who says this?
Possessive relative that’s is geographically wide-spread in the English-speaking world. It has been found in varieties of English spoken in Scotland (Seppänen and Kjellmer 1995; Aitken 1979), Ireland (Harris 1985; Policansky 1982), England (Hudson 1990; Orton and Dieth 1962–1971), the United States (Kurath 1941, Hall 2014, Montgomery and Hall 2004, Pederson et al. 1986), Australia (Newbrook 1992: 22), and Newfoundland and Labrador English (Clarke 2004). Newbrook (1998: 53) notes that it is not found in “new” varieties of English, such as those spoken in many parts of Asia. Our survey data revealed no clear geographical patterns within the U.S., but a regression on demographic categories revealed that younger speakers rate it higher than older speakers.
Singular vs. Plural
Martinez & Wood (2018) found that some speakers accept relative possessive that’s when the object possessed is singular, as in (2a), but not when it is plural, as in (2b). Almost no one had the opposite pattern, accepting singular but not plural.
2) a. The officer that’s keys went missing had to call his boss.
b. %The officers that’s keys went missing had to call their boss.
This effect has been documented in various ways by a number of linguists (Seppänen & Kjellmer 1995; McDaniel et al. 2002; Kayne 2010). Kayne (2010) provides the example in (5a), from Richard Hudson in a discussion on LinguistList (19 Sept 91), where Hudson claims that sentences such as those in (5) are acceptable:
5) a. I’m looking for some pencils that’s leads aren’t broken.
b. I’m looking for a pencil that’s lead isn’t broken.
In the same discussion on LinguistList (20 Sept 91), Lachlan Mackenzie states that in their variety of Scottish English, that’s is only acceptable with a singular antecedent, like the example in (5b).
Subject vs. Object
Possessive relative that’s has been found in both subject and object relative clauses. Subject relative clauses, such as (3a), differ from object relative clauses, such as (3b), by the position within the relative clause that the head noun corresponds to.
3) a. SUBJECT: The kid that’s/whose baseball [ _______SUBJ was signed ] is happy.
b. OBJECT: The kid that’s/whose baseball [ they SUBJ signed _______ OBJ] is happy.
The acceptability of the possessive relative that’s with subject relatives versus object relatives was studied by Stahlke (1976), who tested 35 American students on subject and object relative clauses with that’s, and concluded that subject relatives were preferred. For some speakers, however, that’s is acceptable in object relatives too, as Richard Hudson (LinguistList Sept 91) shows, providing the example:
14) This is the pencil that’s lead you broke.
McDaniel et al (2002) conducted an experiment on the use of that’s in both children and adult speakers of American English. The results showed that adults did not have a preference for that’s with subject versus object relatives; however, children did have a preference for subject relatives with that’s. Martinez & Wood (2018) also studied that’s with subject versus object relative clauses among adult speakers. Among speakers who accept that’s, this study found that subject and object relatives were given similar acceptability ratings: speakers either accepted that’s in both subject and object relative clauses, or they didn't accept it in either.
Animate vs. Inanimate
The third syntactic property to consider is the animacy of the head noun. In (4a), guy is an animate noun, because guy refers to someone who is alive. In (4b), wall is an inanimate noun, because wall refers to something that isn't alive.
4) a. ANIMATE: The guy that’s/whose refrigerator stopped working is back.
b. INANIMATE: The wall that’s/whose paint is peeling is in the kitchen.
The experiments in McDaniel et al (2002: 67) showed that adults did not have a preference for animate or inanimate nouns with that’s, whereas children preferred inanimate nouns. The children's preference for inanimate nouns might be related to the presence of an animacy distinction in the relative pronouns in English who (for animates) vs. which (for inanimates), since which does not have a possessive form (Seppänen & Kjellmer 1995: 397), as illustrated in (17) and (18):
17) a. a boy whose name
b. a book *which's title
c. a book whose title
d. a book the title of which
18) a. a boy that's name
b. a book that's title
(Seppänen & Kjellmer 1995: 397)
Seppänen & Kjellmer (1995) and Aitken (1979), among others, argue that that’s developed from a complementizer that plus the pronoun his in the relative clause subject position. Old English possessive relatives were expressed in this way, and this is also seen in some modern languages, as in (6) (from Seppänen & Kjellmer 1995: 390):
- yesterday-evening'That’s the man whose child died yesterday evening.'
(Scottish National Dictionary (SND), s.v. that)
The idea is that the ’s on that’s started out as a weakened form of his. Then the use of that’s was extended to feminine and neuter nouns as well (Seppänen & Kjellmer 1995: 390). A preference for singular, masculine head nouns could support this idea. However, Seppänen & Kjellmer (1995) and McDaniel et al (2002) do not find a gender distinction in the use of that’s. Moreover, possessive relative that’s is not entirely excluded from relative clauses where the antecedent is a plural noun. Aitken (1979: 105) gives an example of that’s with a plural head that is used in Standard Scottish English:
7) The people that's houses were demolished
Finally, if that’s developed from a weakened masculine pronoun following that, then animate heads might be predicted to be more acceptable. However, adult speakers who accept that’s do not seem to have an animacy distinction (Martinez and Wood 2018). It is possible that that’s did develop from a pronoun, but that it was extended to inanimate nouns quite quickly, perhaps when it was extended to neuter pronouns. But there does not seem to be strong evidence that in favor of this.
Page contributed by Randi Martinez and Jim Wood, February, 2023.
Please cite this page as: Martinez, Randi and Jim Wood. 2023. Relative possessive that's. Yale Grammatical Diversity Project: English in North America. (Available online at http://ygdp.yale.edu/phenomena/. Accessed on YYYY-MM-DD).