Dative Alternation in English

If "He told us a story" is correct, why is it that *"He described us a picture" is not correct?

1.Two possible indirect object positions

For introductory purposes, let us consider the sentence She gave us some new clothes. Gave is a ditransitive verb, in the sense that it takes both a Direct Object and an Indirect Object. The Direct Object is the entity which is (directly) affected by the action of the verb. It is also called theme. In the case of the verb gave, the meaning behind it is that of ‘transfer’. Thus, a Subject points towards where the transfer originated and an Indirect Object expresses where the transfer ends. In a very simplistic view, the Indirect Object benefits from the action of the verb in some way or another hence it is also called goal, or recipient.

In the English language, however, one may find that Sam gave John the book and Sam gave the book to John appear in what seems to be free variation. This feature that the English Indirect Object presents in being able to occupy an after-verb position or an after-preposition position is called dative alternation, dative movement, or dative shift.

2. The importance of studying the dative alternation

In “Explaining English Grammar”, Yule opens the seventh chapter with five very pertinent issues that may arise when teaching the English language: if He told us a story is correct, why is it that *He described us a picture is not correct? Why is *He said me ‘hello’ ungrammatical? If one can say I bet you five dollars, why cannot one say *I bet five dollars to you? Is there a difference between I gave some money to Jack and I gave Jack some money? And why is *I gave a friend it not a valid reply to the question Where’s your bicycle? Leaving aside the fact that there exists a high probability many a teacher of English as a second language may be rather discouraged at these question, it seems that dative alternation has complex, underlying, linguistic issues as a component, the solving and understanding of which may improve teaching English as a second language and our understanding of English syntax, morphophonology, semantics, and pragmatics.

In studying its acquisition and production in children, Pinker, Gropen, Hollander, Goldberg, and Wilson underline that dative alternation could provide a very effective and comprehensive insight into the process of language acquisition and production in general.

The present paper shall briefly look into the two possible dative constructions in English side by side and then it will analyse each of them separately, trying to distinguish meaning differences between the two and see why certain verbs support one construction but not the other. Following such distinctions, I shall present and argue in favour of the three main constraints on alternation / dative alternation criteria proposed by Pinker and Yule.

3. Motivation for the double object construction

Although crosslinguistically it has been noticed that languages which present a “true Dative Case” present a double object construction in complementary distribution with Dative NPs (the to or for Prepositional Phrases in our case).

Levin offers a very simple and solid explanation for the necessity of the double object construction in English. The Indirect Objects, being the recipients, or the entities likely to be given something to (whether concrete or abstract, as in the case with communication verbs), are mostly human. The Direct Objects are the entities given, thus most likely inanimate. The English language needs a structure where the recipient linearly precedes the theme (the double object construction) for “information structure purposes”.

4. Motivation for the after-preposition dative

Personally, I believe that the necessity for the existence of the prepositional Dative is rooted in the fact that the English language has almost no morphology at all and thus no morphological Dative Case. Therefore, there arises an analytical need to build preposition-based construction to substitute for the lack of inflections. A to + Indirect Object form should be no surprise.

One thing to be noted is that this preposition-based Dative has “more language” in its construction than a simple, stand-alone Direct Object. We call this linguistic weight. There is a preference in the English language to place the information with more weight behind it at the end of the sentence (End-Weight Principle), which could easily explain why we find the to + Indirect Object at the end of the sentence.

5. Types of verbs

As can be seen from the fact that He told us a story is correct, but *He described us a picture is ungrammatical, verbs can be categorised into three groups depending on the Dative structure that can accompany them. Trying to find clear criteria by which we can tell exactly which verb will go into which category is one of the goals for the study of Dative Alternation and, although at this stage it appears that the verbs have been inserted into the table below in a completely arbitrary fashion, patterns regarding their meaning and the way they are semantically perceived may be noticed. Thus, there are ditransitive verbs that are only found in double object construction, or only found in prepositional dative construction, or found in both. Largely, the verb groups are:


Group 1 Group 2 Group 3
Both double object and after-preposition dative constructions. After-preposition construction only. Double object construction only.
Bring, build, give, buy, send, cook, teach, get, tell, make Verbs of fulfilling: credit, present, entrust, supply, trust.

Verbs of continuous causation: carry, pull, push, lift, lower.

Others: communicate, construct, describe, create, donate, design, explain, obtain, report, purchase.

Bet, cost, fine, forgive, spare.

6. Establishing differences

6.1 Acquisition and learnability of the Dative

As mentioned above, Pinker et al. have conducted research into the acquisition of the Dative in English children. Their results may help us better understand the two different forms that the English Dative may present itself in. Among their findings, there are two major ideas that deserve consideration. First, it has been noted that neither of the Dative versions consistently emerges first in children’s speech. Contrary to many a grammarian’s opinion, this means that neither version can be considered ‘base’ and the other one derived. Secondly, children do make overgeneralizations about dative alternation in English, meaning they are productive. They use double object constructions, for example, with verbs that they have not heard to be used in that way. As Pinker himself notes, this proves that a strict conservatism hypothesis with regard to the acquisition of dative and dative alternation is wrong. Children do not use a verb solely with the particular dative construction that they have heard the verb being used with.

6.2 Differences in meaning

First of all, a single, quick glance at the table above is sufficient to establish that verbs which belong to the same lexical field are not used in the same constructions.

If one is to consider the double object dative in I (NP0) gave Jack (NP1) the money (NP2) and the prepositional dative in I (NP0) gave the money (NP2) to Jack (NP1), regarding them, Pinker (1989) suggests that their semantic representations differ: from [NP0 causes NP1 to have NP2] in the first case to [NP0 causes NP2 to go to NP1] in the second case.

Krifka, from Humboldt University, also follows Pinker’s idea and believes that, even though the double object and the preposition constructions seem to have the same meaning when a verb may take either one, in truth they do not. He exemplifies this with the obvious difference between send a package to London and send London a package. Krifka concludes that the difference in meaning may make certain verbs compatible with only one of the two structures. I shall refer to this later on as the semantic criterion for dative alternation.

7. The double object construction

There are few features that may be distinguished as belonging to the double object construction alone and again very few features belonging to the verbs that fit in this category alone. Among Krifka’s contributions are his intuition that the DO (double object) imparts a sense of completion of the action that one does not find in the prepositional dative, and that the DO implies that the Indirect Object exists.

Another interesting aspect to note about this construction is that, as Pinker observed, when new verbs enter the language, they sound more natural in the double object form.

Regarding verbs that are only to be found with this form, three broad semantic dimensions may be distinguished for them:

  • Verbs whose action results in some sort of change in the Indirect Object, as in This paper gave me a headache!
  • Verbs such as cost or fine where the Indirect Object is neither a goal, nor a location.

8. The after-preposition construction

According to historians of language, in Old English, the V + dative NP + accusative NP was more used than the one with the accusative NP directly following the verb. Morphological cases started to disappear in Middle English, making way for the V + Indirect Object NP + Direct Object NP that one can see in contemporary English. What happened was that many new verbs entered the English language from French, which marks its goals with the preposition à. The verbs were assimilated alongside the French structure and the preposition was translated as to. The next step was that this borrowing form started to be used with native verbs as well, leading to the prepositional dative form.

As feature pertaining to the after-prepositional structure (PO) grammarians have discerned that:

  • PO is generally preferred when the Indirect Object is a location
  • Verbs that only occur in this structure seem to have a +communication semantic feature.
  • Denoting ‘manner’ of utterance. This may explain why *He said me ‘Hello’ is ungrammatical.

Furthermore, Yule notices that, by placing the Indirect Object after a preposition, and, thus, farther away from the verb, the speaker creates greater linguistic distance. In the English language, linguistic distance is directly proportional to conceptual distance, meaning that the closer an object is placed to the verb, the closer the connection it has with it. If the speakers perceives there is a looser connection between the action of the verb and the recipient, the after-prepositional structure will be used.

9. The morphophonological criterion

Although observation and history are accurate tools in helping categorize which verbs and which circumstances require which of the two structures, the real linguistic interest in the study of dative alternation is the formulation of constraints on or criteria for alternation: why do some verbs or circumstances require one form and not the other?

Grammarians agree that there are three such major criteria for dative alternation in English.

The first of these is the morphophonological one: double object verbs belong to the native stem class rather than the Latinate class. A phonological criterion is also at play here since children that have started to acquire English as their native tongue have no knowledge of etymology, but they rely unconsciously on the phonological distinction that stem verbs are generally monosyllabic or polysyllabic with initial stress. It is interesting to note that Latinate verbs which do behave like native verbs also have initial stress.

10. The Semantic criterion

The semantic constraint on alternation states that the Indirect Object of a double object construction must be both a possessor and a goal if the structure alternates with a to + dative or a possessor and a beneficiary if the structure alternates with a for + dative.

Pinker offers the following example: John gave Mary a car is a sentence containing a verb that would alternate with to + dative (John gave a car to Mary) so, according to the semantic criterion, in order for the double object construction to be correct, the Indirect Object (Mary) would have to be both a goal, which it is by its nature of being an Indirect Object for the verb give, and a possessor, which it is because the entity Mary shall possess said car due to the action of the verb. The semantic criterion is verified, so the construction is correct. However, in *John washed Mary a car (which again would alternate with a to + dative construction), the Indirect Object Mary, although it may be a goal, is not necessarily the possessor of the car. The semantic criterion is not met, hence this double object structure is ungrammatical. The same reason accounts for the ungrammaticality in the example given by Bresnan 1978, *I sent the border a package. The border is inanimate and therefore it cannot possess anything.

There are certain verbs that meet the semantic criterion, but are not found in the double object construction. These exceptions include: say, pulled, shouted, credited, chose etc.

11. The pragmatic criterion

There is also another consideration by which one may choose between one type of dative structure over the other: when the theme is known and the goal is new information, the prepositional dative is preferred; when the theme is new, the double object dative is more appropriate. This means that material containing new information tends to present itself at the end. So, for example, if one is to look at She gave John money versus She gave money to John, although it has already been shown above how Pinker considers the two are different from the point of view of semantic representation (and subsequently meaning), one may also infer that in the first example, “money” is the new piece of information in the discourse, while in the latter example, the fact that said money was given to a “John” is the new information.

12. Errors children make with dative shift

One may argue that the constraints on dative alternation presented above may not hold as true criteria because of the fact that children acquiring the English language commit errors regarding dative shift all the time. The research of Pinker et al. on acquisition and learnability of the dative alternation, however, showed that the errors children make are due to simple substitutions of verbs and not overextensions of dative rules.

13. Conclusions

I have very briefly highlighted the most important aspects regarding when dative alternation happens, as well as why it happens, according to the research of various grammarians whose theories I fully support. The study of dative shift is relevant for the purpose of explaining why certain syntactical paradigms cannot be found with what seem to be almost identical elements as well as for the purpose of highlighting differences in meaning and/or focus in pseudo-synonymic structures. Besides giving categories of verbs and discourse circumstances from the point of view of their behaviour regarding the position of the Indirect Object in the English sentence, such a study also tries to reveal the underlying mechanisms which prove that grammar is not (entirely) an arbitrary matter of convention. Even more, the study of how children tackle dative movement has proven to be very useful in the understanding of language acquisition in general, concerning overextension of grammatical rules and other such matters.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s