Academic journal article Memory & Cognition

Word Order Variation in Spatial Descriptions with Adverbs

Academic journal article Memory & Cognition

Word Order Variation in Spatial Descriptions with Adverbs

Article excerpt

Previous research has shown that in a three-term spatial reasoning task, the second premise of a German premise pair is especially easy to comprehend if (1) the prepositional object rather than the grammatical subject denotes the given entity, and if (2) the term denoting the given entity precedes the term denoting the new entity. Accordingly, the second premise is easiest to comprehend with noncanonical word order-that is, with the prepositional object in preverbal position denoting the given entity (e.g., To the right of the given object is the new subject). This finding is explained in terms of contextual licensing of noncanonical word order. Here, we discuss and tested two alternative accounts of contextual licensing, given-new and partially ordered set relations (Poset). The given-new account claims that noncanonical word order is licensed by the term denoting the given entity preceding the term denoting the new entity. On the Poset account, noncanonical word order is licensed if the preverbal constituent introduces a new entity that stands in a transitive, irreflexive, and asymmetric relation to a given entity. Comprehension times for second premises with spatial adverbs in four different word orders support both accounts of contextual licensing; Poset licensing was stronger than given-new licensing.

When people read a spatial relational statement, such as The snake is to the left of the deer, and when such a statement is continued by a second one, such as The donkey is to the right of the deer, they are capable of making a spatial inference-in the present instance, that the snake is to the left of the donkey. For a valid conclusion, the two statements that constitute the premise pair must meet two conditions. First, the premises must have exactly one term in common (the middle term)-the deer in the example just given. second, the description as a whole must be determinate insofar as either premise must relate two adjacent entities of the layout to one another (i.e., the middle term must denote the inner entity of a linear layout). According to the mental model theory of reasoning (Johnson-Laird, 1983), people arrive at the conclusion by initially constructing a mental model of the layout described by the first premise. The second premise is then interpreted in the context of the initial mental model. The middle term in the second premise-the deer-denotes the given entity already included within the initial model, whereas the other term in the second premise (the end term)-the donkey-denotes a new entity not mentioned previously. Processing the second premise requires the integration of the new entity into the initial model. This process of premise integration constitutes the core of the reasoning process (see Huttenlocher & Higgins, 1971). Since the middle term anaphorically refers to a previously introduced entity, premise integration in spatial reasoning exemplifies an instance of the processing of anaphoric relations in general.

For ease of exposition and in order to keep apart the different levels of linguistic representation that are involved here, we will begin with some terminological clarifications for a sentence of the form The x is to the left of the y. On the syntactic level, we differentiate between the external argument and the internal argument of the preposition to the left of. The external argument appears as the grammatical subject, the x, of the sentence. The internal argument appears as the noun phrase they, which is part of the prepositional phrase. Below we will use the term (grammatical) subject to designate the external argument of the preposition and the term (prepositional) object to designate the internal argument of the preposition. Note that in The x is to the left of the y, the subject precedes the verb (the copula to be). The grammatical subject in preverbal position identifies the word order as canonical, a point that we will return to below.

On the semantic level, we differentiate between the referent (i. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.