The Sleeping Child

Chomsky 1957 discusses the following paradigm contrasting adjectives like interesting and present participles like sleeping (see pages 73-75):

(1) a. The book is interesting.

b. The child is sleeping.

(2) a. The book is very interesting.

b. *The child is very sleeping.

(3) a. The book seems interesting.

b. *The child seems sleeping.

(4) a. the interesting book

b. the sleeping child

(2) shows that very modifies adjectives but not present participles.  Similarly, although seems can take an adjectival complement (as in (3a)), it cannot take a present participle complement (as in (3b)). The examples in (4) show clearly that the prenominal position is not restricted to adjectival modifiers. Rather, present participles are also allowed.

Chomsky (1957: 72) argues that the word order in (4) for both adjectives and present participles is derived by the transformation T which converts “the boy is tall” into “the tall boy”. I will put aside adjectives for the moment to focus on the present participle sleeping. 

The example in (5) shows that even in prenominal position, sleeping is acting as a present participle, not an adjective. 

(5) a. the very interesting book

b. *the very sleeping child

We can add to this paradigm the fact that sleeping can be modified by manner adverbs, which normally appear with eventive verbs:

(6) a. The child is quietly sleeping.

b. The child is restlessly sleeping.

(7) a. the quietly sleeping child

b. the restlessly sleeping child

These data strongly suggest that sleeping in the prenominal position is a present participle, capable of being modified by an adverb, not an adjective. It follows that (4b) is a prenominal relative clause construction:

(6) In English, prenominal present participles are prenominal relative clauses.

Following Kayne 1994, I adopt the head-raising analysis of relative clauses. Applied to the example at hand, we have the following derivation. I put aside the issue of whether or not there are further functional projections (such as TP) in the relative clause):

(7) a. Merge(child, sleeping)  = [vP child sleeping]

b. Merge(C, vP)   = [ C [vP child sleeping]]

c. Merge(child, CP)   = [CP child [ C [vP <child> sleeping]]]

d. Merge(the, CP)   = [DP the [CP child [ C [vP <child> sleeping]]]]

This derivation yields a postnominal relative clause, not a prenominal relative clause. Postnominal relative clauses based on participles are acceptable when made a bit heavier by the presence of an adverb (see 8b)). I will not pursue this issue here.

(8) a. ??The child sleeping started to cry.

b. The child sleeping over there started to cry.

To get the prenominal word order, the structure will have to be inverted. I propose using the linker phrase LkP, which allows for a change in word order (see Collins 2019). The first three steps are identical to (7):

(8) a. Merge(Lk, CP) = [LKP Lk [CP child [C [vP <child> sleeping]]]]

b. Merge(vP, LkP) = [CP [vP <child> sleeping] [LK’ LK [CP child [ C vP]]]]

c. Merge(the, LkP) =       [DP the [[vP <child> sleeping] [LK’ LK [CP child [ C vP]]]]]

This derivation yields the right word order: the sleeping child. 

The derivation of prenominal present participial relative clauses in English is largely the same that Kayne (1994:98) proposed for passive participles, as in (9). I will not pursue the rather minor differences in the analyses here.

(9) the recently sent book

Furthermore, the fact that participles (present and passive) are derivations like that in (8) raises the possibility that adjectives in general also have relative clause derivations (on which see Kayne 1994: 8.4)

Acknowledgments: Thanks to Gary Thoms for discussing the topic of this paper with me.

References:

Chomsky, Noam. 1957. Syntactic Structures. Mouton, The Hague.

Collins, Chris. 2019. The Linker in the Khoisan Languages. Oxford University Press, Oxford.

Kayne, Richard. 1994. The Antisymmetry of Syntax. MIT Press, Cambridge.