LINGUIST List 15.2526

Fri Sep 10 2004

Disc: New: Review: Evans, Vyvyan (2004)

Editor for this issue: Marie Klopfenstein <>


  1. Vyv Evans, Response to review of "The structure of time" (15.2430)

Message 1: Response to review of "The structure of time" (15.2430)

Date: Fri, 10 Sep 2004 09:53:29 +0100
From: Vyv Evans <>
Subject: Response to review of "The structure of time" (15.2430)

I am grateful to Thora Tenbrink for bringing attention to my recent
book, "The structure of time", in her review posted on Linguist 30th
of August 2004, issue 15.2430. Her review was generally positive and
I welcome her conclusion that the book is "thought-provoking and
inspiring. It is a valuable interdisciplinary source for insight in
several domains, including at least lexical semantics, conceptual
metaphor theory, and cognitive science in the area of time." However,
there are three key misunderstandings apparent, which relate to areas
of theory, analysis and methodology. These serve to misrepresent
central aspects of the book. I therefore address these below.

I turn first to a misrepresentation in the review which goes to the
very heart of the research project which this book is part of. In
discussing the chapter addressing what I term the Agentive Sense for
time, the reviewer says the following: "Evans claims that the Agentive
Sense is not metaphorical in nature. This entails that utterances like
"Time is the great physician" are supposed to be interpreted
literally, mirroring the speaker's concept of an agent that produces
similar effects as other y"inds of agents." This represents a
serious misunderstanding of what the book is arguing. The basic
theoretical premise of the book is that how we conceptualize time, as
encoded in language, constitutes two levels of organization. First,
there is the notion of a lexical concept, a level of representation
that relates to discrete units of language such as the word 'time',
which is polysemous. The Agentive Sense constitutes one such distinct
lexical concept conventionally associated with the English lexeme
'time'. There is another level, the level of cognitive models, which
are larger-scale representations involving integrations of distinct
lexical concepts for time. These constitute much more complex
representations for time.

A central axiom in cognitive linguistic theory is that there are
patterns of imagery associated with (or constitutive of) mental
representations, and these are highly conventional. The influential
framework of conceptual metaphor theory, associated in particular with
Lakoff and Johnson (e.g., 1980, 1999), has argued for quite abstract
mappings to account for this, termed conceptual metaphors. In recent
work, since at least Lakoff and Turner (1989) which argued for a
"generic level" in metaphoric mappings, and more recently in arguing
for "primary metaphors" in their 1999 book, "Philosophy in the flesh",
Lakoff, Johnson, and others, have argued that the "foundational" level
of metaphor is highly schematic.

However, this move towards more schematic foundational representations
constitutes an impulse which can be construed as moving in opposition
to another central axiom of cognitive linguistic theory, which is that
mental representation, as encoded in language, is highly redundant,
maximalist and specific. That is, models of conceptual structure, as
encoded in language, should take a "bottom-up" perspective.

The approach taken in "The structure of time" can be described as
"bottom up" in this sense. Rather than assuming that lexical concepts
receive their conventional patterns of imagery due to highly schematic
conceptual metaphors, e.g., "time is motion", with two celebrated
variants, the so-called "moving time" and "moving ego" mappings, the
argument is that conventional patterns of imagery are licensed by
antecedent lexical concepts. These are much more specific, in terms
of pre-existing structure, than is normally assumed by metaphor
scholars for target concepts/domains. That is, I argue for a range of
distinct lexical concepts associated with the lexeme 'time', each of
which has distinct patterns of conventional imagery.

Accordingly, it is quite wrong to say that I argue that the Agentive
Sense, or any other sense associated with time is not metaphoric, in
the sense of lacking conventional patterns of imagery. Indeed, I use
the term "concept elaboration", rather than "metaphor", although I'm
attempting to account for the same range of phenomena. Indeed, a
significant proportion of the book, particularly the sections in the
chapters in part II which address concept elaboration represent an
attempt to describe and account for the nature of the imagery
associated with lexical concepts for time. Moreover, in chapter 4 I
introduce two mechanisms, experiential correlation and perceptual
resemblance, due to the work of metaphor scholar (Grady, e.g., 1999),
which I employ in order to account for these patterns.

In addition, at no stage do I claim that patterns of imagery
associated with the Agentive Sense such as "time is the great
physician" are understood literally. To suggest that such a view is
entailed, as the reviewer does, constitutes a serious misreading of
the chapter in question, and the argument developed in the book.
After all, the main premise of the book is that although the human
experience of time is phenomenologically real, it is nevertheless, at
base, a subjective experience, which can arguably be traced back to
perceptual mechanisms and processes. What I am arguing is actually
quite close to what the reviewer suggests I should be arguing.
Namely, expressions such as "time is the great physician" are
conventional patterns of imagery associated with a "complex" lexical
concept (I use the term "secondary lexical concept"), what I call the
Agentive Sense.

The point I make in chapter 12, which may have given rise to the
misreading, is that just as highly schematic metaphoric mappings (in
the sense of Lakoff and Johnson's recent work) may arise from the
integration of more specific lexical concepts and their conventional
patterns of imagery (see my discussion in part III of the book), the
proliferation of "specific level" metaphors argued for in Lakoff and
Turner (1989) can also be reconceptualized. Rather than positing a
multiplicity of distinct metaphors for time which relate to notions of
personification, by assuming a single lexical concept, an Agentive
Sense, which licenses a range of conventional patterns of elaboration,
we constrain (and better account for) the linguistic data. That is,
just as highly abstract mappings may be suspect, so too an unwarranted
proliferation of distinct "agentive" metaphors may also be suspect.
Indeed, this notion of appropriately constraining conceptual
structure, as encoded in language, and yet recognizing its flexibility
and creativity is one of the hallmarks of the Principled Polysemy
framework initially developed in joint work with Andrea Tyler and
extended here.

A second key misinterpretation relates to the discussion in chapter 18
of what is termed the 'temporal sequence model'. The reviewer
criticises this account for failing to appreciate that such a model of
time does not rely on the participation of an ego/experiencer. That
is, there is no (human) deictic centre. However, the whole pointy of
my account is to stress that what is at stake is not so much the
presence or absence of an ego, but rather that the reference point or
deictic centre in this particular model is time-based rather than
ego-based. Put another way, the experiencer is not 'on-stage' in this
model, while in the two other models I deal with, the ego/experiencer
is 'on-stage'. That is the ego/experiencer provides his/her own
temporal "location" as a reference point to "anchor" the system in
"ego-based" models for time. Indeed, in recognizing this, I follow
work by Traugott (1978), who I discuss, and more recent work by Kevin
Moore (2000) in his research on time metaphors in Wolof.

My account of the very interesting difference between English and
Hausa reflects the fact that Hausa and English, in certain
circumstances, encode similar temporal scenes involving time-based (as
opposed to ego-based) reference in strikingly different ways. This
relates to similarities in the way Hausa encodes both spatial frames
of reference and temporal frames of reference (which is essentially
what cognitive models for time are; see Evans To appear). In Hausa,
spatial scenes can be encoded using spatial language which evokes the
notion of an 'off-stage' experiencer (or at least an off-stage
viewpoint). This viewpoint differs from that of English. Thus, my
use of in-tandem alignment is not an attempt to bring in an ego into
the temporal sequence model as a way of "anchoring" temporal reference
in this model, as I make explcit, but rather constitutes an attempt to
account for the strikingly different patterns in the language used by
these two languages for describing similar models of temporal

A third misrepresentation relates to the reviewer's claim that there
are no "hard criteria" for establishing the categories of distinct
lexical concepts I argue for. In particular, the reviewer claims that
"no specific caution is taken [sic] that the diverse readings do not
stem from contributions of the other lexical items in the examples,
rather than the underlying concepts of 'time'." In this, the reviewer
claims, I fall short of my previous work (Tyler and Evans 2003) which
did give such an admonition, with respect to classifying lexical
concepts associated with English prepositions. First, let me point
out that I posit three principles in "The structure of time" which
relate to meaning, grammar and patterns of imagery. These
collectively serve to identify distinct lexical concepts. While these
criteria might not be the right ones, or may give an artifical view of
lexical concepts as discontinuous, when relationships between lexical
concepts might be better represented as more of a continuum (as I
acknowledge in the book), it is patently wrong to say there are no
criteria for establishing the categories I argue for. The central
methodological claim of the book is that (cognitive) linguistic theory
in general, and (cognitive) lexical semantics in particular, do
require "decision principles" (as Sandra (1998) calls them), in order
to make "cuts" in conceptual structure, based on linguistic data. In
the book, the examples I provide are all classified on the basis of
analysis using the critera I put forward. For instance, the Event and
Moment Senses constitute distinct senses based on the criteria I
advance, not my intuitions, as I make clear.

Moreover, in terms of methodology this book improves on the earlier
2003 book. In that book, while Tyler and I admitted that other
lexical items contribute to the meaning of a particular word (the
notion of "sentential context"), this work explictly builds sentential
context into the criteria by positing a grammatical criterion and what
I call a "concept elaboration" criterion for sense disambiguation.
That is, grammatical context, and elaborative context (for example,
the imagery associated with the time concept in question) are taken
into account when deciding the particular category that any given
usage of 'time' should be classified as an instance of.

Finally, let me address a perhaps more minor point from the review
that might cause potential confusion. The reviewer describes it as
unfortuate that I only use two languages (Aymara and Hausa) in
addition to English in my analysis. This might suggest to some
potential readers that the book is attempting a cross-linguistic
investigation and thus, in mainly employing English I'm being
negligent. This is not the case. The study, particuarly in part II,
is primarily a lexical semantic analysis of the English lexeme 'time',
as I make clear. The very brief discussion of Aymara and Hausa in
part III is there simply to make the point that larger-scale cognitive
models for time are culture-specfic. Aymara and Hausa have similar
cognitive models for time to English, but structure and elaborate
these models in slightly different ways which result in strikingly
different conceptualization and lexicalization patterns. Indeed, as I
acknowledge, apart from a few reasonably well-studied languages,
mainly European, how other languages structure time is a vast
uncharted area which urgently requires the attention of linguists.
"The structure of time" represents, in part, an attempt to provide a
methodology that might provide a framework for such an undertaking.

Evans, Vyvyan. To appear. How we conceptualize time: language,
meaning and temporal cognition. Essays in Arts and Sciences,
vol. XXXIII, Oct. 2004.

Grady, Joseph. 1999. A typology of motivation for conceptual
metaphor: Correlation vs. resemblance. In R. Gibbs and G. Steen
(eds.). Metaphor in cognitive linguistics (pp. 79-100). John

Lakoff, George and Mark Johnson. 1980. Metaphors we live by.
Chicago UP.

Lakoff, George and Mark Johnson. 1999. Philsophy in the flesh: The
embodied mind and its challenge to western thought. Basic Books.

Lakoff, Geoge and Mark Turner. 1989. More than cool reason: A field
guide to poetic metaphor. Chicago UP.

Moore, Kevin Ezra. 2000. Spatial experience and temporal metaphors in
Wolof. Unpublished doctoral thesis, Linguistics Dept., UC Berkeley.

Sandra, Dominiek. 1998. What linguists can and can't tell us about
the mind: A reply to Croft. Cognitive Linguistics, 9, 4, 361-378.

Traugott, Elizabeth Closs. 1978. On the expression of
spatio-temporal relations in language. In J. Greenberg (ed.).
Universals of human language. (pp. 369-400). Stanford UP.

Tyler, Andrea and Vyvyan Evans. 2003. The semantics of English
prepositions: Spatial scenes, embodied meaning and cognition. CUP

Vyv Evans
Department of Linguistics and English Language
University of Sussex
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue