LINGUIST List 15.296

Tue Jan 27 2004

Review: Semantics: Tyler & Evans (2003)

Editor for this issue: Naomi Ogasawara <>

What follows is a review or discussion note contributed to our Book Discussion Forum. We expect discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for review." Then contact Sheila Dooley Collberg at


  1. Thora Tenbrink, The Semantics of English Prepositions

Message 1: The Semantics of English Prepositions

Date: Tue, 27 Jan 2004 10:24:43 -0500 (EST)
From: Thora Tenbrink <>
Subject: The Semantics of English Prepositions

Tyler, Andrea & Vyvyan Evans (2003) The Semantics of English
Prepositions: Spatial Sciences, Embodied Meaning, and
Cognition. Cambridge University Press.

Announced at

Thora Tenbrink, University of Bremen, Germany


This book is a comprehensive account of English spatial prepositions
in the cognitive linguistics tradition. Pursuing a ''principled
polysemy'' approach, the authors present an analysis of the semantic
networks of altogether 15 spatial prepositions, giving a detailed case
study of 'over' and sketching the basic features of the other
particles. Many of the presented insights are well-founded in the
literature -- major sources are Langacker (e.g., 1987, 1991), Talmy
(e.g., 2000), and Grady (1997) -- while others are original. Central
to the book is the methodology developed to identify primary and
secondary senses associated with each preposition.

While it is the authors' conviction that all other senses of a
preposition are derived from an originally and primarily spatial
sense, the relationship is nevertheless not viewed as metaphorical as
in various previous approaches, but as experiential. In their view,
experiencing regular and motivated correlations of events and
scenarios in the real world motivates the integration of such
regularities into language. In this way, 'up' became associated with
'more', since it is a regular experience in spatial scenarios that
larger amounts correlate with increased vertical extension (e.g.,
liquid in a glass). Together with pragmatic strengthening it may then
happen that concepts that were originally simply associated with other
concepts in the usage of a certain preposition, come to be expressed
by that preposition even without the presence of the original
concept. Thus, in 'The prices have gone up' there is no longer a
spatial sense of vertical direction involved. Basically, Tyler and
Evans' approach is in line with previous analyses in the cognitive
linguistic tradition, distinguished mainly in their methodology that
is designed to steer away from the traditional, rather vague
conception of underlying metaphorical relationships, towards a more
systematic investigation of how such cognitive dependencies may be


Chapter 1. The nature of meaning.

The first chapter serves as an introduction presenting the main
aspects of the authors' theoretical views: Spatial prepositions have
developed their semantic potential on a transparent experiential
basis. The meanings associated with each particle are related in
systematic and motivated ways (underscoring the adopted principled
polysemy in contrast to homonymy or monosemy approaches). As a general
trait, linguistic expressions are underdetermined; they serve as
prompts for meaning construction, which is a conceptual process. In
other words, language does not directly refer to the real world but to
human conceptualizations of it. The human conceptual system, in turn,
is a product of how we experience and interact with the real world. In
this way, linguistic expressions and their usage are shaped by our
experience. In comprehension, the selection of a particular meaning
out of the available meaning potential is determined by conceptual
integration in the given context.

Chapter 2. Embodied meaning and spatial experience

In this chapter, the authors develop their ideas as to how the
functional nature of spatial scenes give rise to correlated
non-spatial inferences, a process they term ''experiential
correlation'' (cf. Grady 1997). For instance, spatial relations such
as containment are interconnected with experiences such as being
protected, being constrained and delimited, etc., which constitute
functional rather than spatial relations. Similarly, perceptual
resemblances can lead to perceived shared characteristics, as
exemplified by 'She's just a twig.' The authors propose that
underlying such a sentence is not metaphorical transfer but rather a
''process which relies on conscious human perceptual organization'',
which, just as experiential correlation, leads to the
conceptualization of one kind of entity in terms of another.

Chapter 3. Towards a model of polysemy networks: spatial scenes and

In the third chapter, the authors warn that the polysemy approach can
lead to the fallacy of exaggerating the number of distinct senses
associated with a preposition. To avoid this problem, they propose a
methodology for determining a particle's ''primary sense'' and filter
out further distinct senses that are not derivable from the
context. For the primary sense, the following criteria are put

1. earliest attested meaning,
2. predominance in the semantic network (the unique spatial
 configuration that is involved in the majority of the distinct senses
 found in the network),
3. use in composite forms,
4. relations to other spatial particles,
5. grammatical predictions.

To determine further distinct senses, a candidate sense must
1. contain additional meaning not apparent in any other senses associated
 with a particular form, and
2. there must be instances of the sense that are context independent (the
 distinct sense could not be inferred from another sense and the context
 in which it occurs).
For the primary sense, a spatial proto-scene can be depicted that is
an abstraction of real-world spatial scenes represented by the
particle. Furthermore, each primary sense is associated with a
functional element (such as ''containment'' for 'in').

Chapter 4. The case of over

In this chapter, Tyler and Evans provide a detailed analysis of 'over'
on the basis of their proposed methodology to illustrate how it can be
applied. They offer a spatial proto-scene for 'over' in which the
trajector is higher than but within potential contact of the landmark.
The associated functional element is that landmark and trajector are
conceptualized as being within the sphere of influence of each ofter.
Contrary to previous analyses, the authors claim that 'over' does not
have an associated ''Path'' sense, neither in its primary sense nor in
its other distinct senses. They go to some length in explaining how,
according to their analysis, all path interpretations in 'over'
sentences can be derived from the context. Furthermore, the proposed
semantic network of 'over' is presented in full, giving evidence as to
how each distinct sense - such as the ''Above-and-beyond'', the
''Completion'' and the ''Transfer'' senses, among others - may have
been derived from the primary sense.

Chapter 5. The vertical axis

The notion of a ''contrast set'' is introduced. The authors point out
that lexemes often do not constitute oppositional pairs, but are
simultaneously dependent on and independent of other lexemes. Thus,
the contrast set 'over', 'above', 'under' and 'below' divides up the
vertical axis into four distinct spatial locations. But the functional
elements associated with each particle are dependent on the
implications of each specific spatial proto-scene and are therefore
not complementary in the same way as the spatial meanings. With regard
to the spatial arrangement, a (subjective) sense of proximity is
claimed to be associated with 'over' and 'under', as contrasted with
distality in 'above' and 'below'. As a consequence, the latter
prepositions do not allow for contact, and they are not conceptualized
as involving possible influence between trajector and landmark. But
the semantic network of 'under' does not mirror that of 'over' because
entities that are lower in verticality are often less accessible;
therefore, there are fewer experiential correlations associated with
'under' than with 'over'.

Chapter 6. Spatial particles of orientation

Chapter 6 deals with particles whose semantics - according to the
authors' analysis - involves orientation of either trajector ('up',
''down'', ''to'', ''for'') or landmark (''in front of'', ''before'',
'behind', ''after''). A clear distinction between orientation, path,
and motion is made; these three concepts are often conflated in other
approaches since they often occur together. Before/after differ from
in front of/ behind in that they involve a sense of sequence: the
trajector is construed as leading/following the landmark.

Chapter 7. The bounded LM

In this chapter, prepositions are analysed that are sensitive to
boundedness on the part of the landmark: 'in', 'into', 'out', 'out of'
and 'through'. Bounded entities are defined as those that possess an
interior, a boundary and an exterior. Often, they are associated with
the notion of containment, which itself entails several functional
consequences which are reflected in the various senses of 'in' (and
partly 'out' as well).


Altogether, the book gives an impressive insight into the semantic
networks of spatial prepositions, showing clearly and (for the most
part) convincingly how new meanings develop from established ones on
the basis of experiential correlations. Accounting for such processes
in language without simply relying on the often too vague concept of
''metaphorical extension'' which has frequently been claimed to be
responsible for the usage of similar forms in different domains is a
major achievement. Furthermore, the notion of ''contrast set'' is
very useful in accounting for the relationships of prepositions to
each other, which are often not sufficiently explained by opposing
features. Thus, the basic approach is both innovative and convincing
and will provide thought-provoking impulses for a broad readership.

In the following, I will point to some specific weaknesses that do not
diminish the value of the book's achievements as a whole. The book is
readable and well-edited, providing a clear general structure and a
detailed keyword index. However, one major shortcoming on this scale
is the pervasive usage of large footnotes - sometimes extending over
almost one page. The footnotes often do not relate directly to the
main text but digress considerably from the current train of
thought. At times, issues dealt with in footnotes are taken up later
in the main text, providing no new information: in those cases, the
earlier footnote should either be left out or simply refer to the
later main text. Another problem related to the book's structure
concerns the fact that, in spite of the authors' constant reference to
their own proposed methodology, they do not present a comprehensive
overview of the proposed criteria for determining the primary and
further senses, for example, in the form of a list with detailed and
specific information with regard to how the criteria should be
applied, which could be used for reference. Furthermore, in spite of
the fact that they refer to ''semantic networks'', which could easily
be visualized, they only provide a graphical depiction of the network
of 'over', but not for the others. Since diagrammatic representations
are often amazingly informative with regard to the implied relations
(which is obvious through the depictions of proto-scenes provided
throughout), the authors thus abstain from a valuable and feasible way
of conveying insights about the analyses of the other 14 prepositions
that could not be presented in detail.

In line with previous approaches in the cognitive linguistics
tradition, the authors mostly rely on their own intuitions regarding
the potential usage of the prepositions under analysis. They provide
convincing naturalistic examples that seem to cover the variations of
usage fairly well. However, as the authors note themselves in the
conclusion, without a systematic corpus-based analysis this remains
speculative. Considering the authors' proposal that one important
ingredient of the proposed methodology is the criterion of
predominance, this is a major drawback. The authors' proposed
methodology is altogether fairly convincing. However, it is not
always transparent in how far it is actually put to use in their
analysis. This may in part be due to the fact that a comprehensive
summary of the methodology is missing, in part to the fact that they
do not apply the criteria systematically one after the other, and in
part to the fact that some of the proposed criteria (e.g., the
''predominance'' criterion and those proposed for determining
additional distinct senses) still very much depend on subjective
assessment. Thus, the reasons for the decisions made are sometimes not
traceable and therefore still open to discussion.

One major point of discussion would be the authors' claims about
prepositions that are not predominantly spatial in current usage, such
as 'before' or 'of'. In the analysis of 'before' they propose a
spatial ''In-advance-of'' sense involving an ''in tandem''
configuration for the proto-scene, from which a temporal sense is
derived. While the view that 'before' should be regarded as originally
spatial is quite prevalent in the literature, it does not become clear
from the present analysis how this could be proved (nor can this view
be regarded as self-evidently true since it is subject to much
controversy). The authors themselves acknowledge that 'before' in
current usage is mostly regarded as a predominantly temporal
preposition, and that the etymological aspect in this case does not
suffice for determining the proto-scene. Even the earliest uses of
'before' were overwhelmingly sequential and temporal - two facets of
meaning both involving time. Thus, their proposal of a spatial primary
sense of before seems to be due to their theoretical orientation
towards spatial scenes, rather than objective criteria. That
temporality in ordinary 'before'-sentences is crucial perhaps becomes
most evident when contrasting them with one of the few occurrences of
locational 'before':

(1) They knelt before the Queen.
(2) They knelt before the Queen did.

(1) can be characterized by a functional locational interpretation, as
outlined by Tyler and Evans. But crucially, 'before' does not
temporally divide the scene into an earlier and a later event, as is
the case in (2), where two events happen at two different times. No
specific orientation (such as an ''in tandem'' configuration) of the
participants is needed. Thus, the two senses of 'before' differ
fundamentally with regard to the temporality they are capable of
expressing, contrary to the authors' claim that there is an underlying
spatial proto-scene which corresponds closely to both. It should
further be noted that Tyler and Evans (like many other authors dealing
with these terms) do not comment on the striking fact that 'before'
and 'after', unlike most other terms under analysis, can also be used
as a conjunction - often with very similar semantics (though regularly
with different implications) as the prepositional form, as in: 'They
arrived before Bill (did).'

In the case of 'of', they state that ''it appears to be one of only a
few spatial particles, which, in synchronic terms, seems to have
become largely dissociated from its spatial origin''. In other words:
What makes the particle ''spatial'' for the authors is its etymology
rather than any of the other objective criteria proposed to determine
a particle's spatial proto-scene.

Considering these problems, one wonders whether what is stated in the
book's cover abstract really reflects the authors' stance, namely,
that ''all English prepositions originally coded spatial relations''
(a statement which accords with the book's title, which strangely does
not restrict the book to spatial prepositions). If that were the case,
explanations for prepositions like 'during', 'despite', 'since', etc.
must be regarded as missing in the book. Quite naturally, however, the
authors do not attempt to provide spatial proto-scenes for such
uncontroversially non-spatial prepositions (although, notably, it is
surely not impossible to find spatial representations for abstract
relationships of other kinds denoted by prepositions).

Altogether, the book is highly recommended to all researchers
interested in the semantics and usage of spatial expressions as well
as the underlying cognitive processes. It offers new insights and a
promising general approach to the treatment of semantic networks.


Grady, Joseph. 1997. Foundations of meaning: Primary metaphors and
primary scenes. PhD dissertation, Dept. of Linguistics, UC Berkeley.

Langacker, Ronald W. 1987, 1991. Foundations of cognitive grammar. Vol
I (1987): Theoretical Prerequisites. Vol II (1991): Descriptive
Application. Stanford, CA: Stanford University Press.

Talmy, Leonard. 2000. Toward a Cognitive Semantics, 2 vols. Cambridge,
MA: MIT Press.


Thora Tenbrink is a research assistant in the DFG Collaborative
Research Center SFB/TR8 "Spatial Cognition: Reasoning, Action,
Interaction" (Bremen & Freiburg, Germany). Her dissertation project
deals with the question how objects and events are localised relative
to other objects and events using spatial and temporal expressions in
natural discourse. Previous work has dealt with discourse relations
and information structure, presuppositions and non-temporal
implications of temporal connectives, especially 'before' and
'after'. Her current focus is on empirical research on spatial
reference systems in human-robot interaction.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue