Review of Annual Review of Cognitive Linguistics
|Date: Wed, 04 Aug 2004 00:37:37 +0200
From: Suzie Bartsch
Subject: Annual Review of Cognitive Linguistics, Volume 1
EDITOR: Ruiz de Mendoza Ibáñez, Francisco José
TITLE: Annual Review of Cognitive Linguistics
SUBTITLE: Volume 1
PUBLISHER: John Benjamins Publishing Company
Suzie Bartsch, unaffiliated
This first issue (iv+288 pages) of the ''Annual Review of
Cognitive Linguistics'' (henceforth ARCL issue) is a collection
of 11 fascinating papers, an interesting interview with George
Lakoff, and 2 enlightening book reviews, including
methodological (Gries), empirical (Gries, Caballero, Ortigosa
Pastor, Porto Requejo, Vallès, Sánchez García, Nerlich &
Chamizo Domínguez) and 'pseudo-empirical' or 'semi-empirical'
(Ibarretxe-Antuñano, Hamawand, Paradis), as well as
theoretical (Paradis, Kravchenko's paper and the review of his
book by Shakovsky) approaches to Cognitive Linguistics (CL).
The contributions address issues which proved to be, or are
becoming, crucial within CL, such as:
- metaphor theory (Caballero, Porto Requejo, Sánchez García,
Lakoff's interview by Sánchez),
- prototype theory and conceptual networks (Gries, Ibarretxe-
Antuñano, Hamawand, Porto Requejo, Vallès),
- polysemy (Hamawand, Nerlich & Chamizo Domínguez, Paradis),
- metonymy (review of Ruiz de Mendoza Ibáñez et al. 2002 by Panther),
- Neural Theory of Language (NTL) (Lakoff's interview),
- image schemas (Ibarretxe-Antuñano, Lakoff's interview),
- constructionist approaches (Gries, Ibarretxe-Antuñano,
Hamawand, Ortigosa Pastor, Lakoff's interview),
- lexical creativity (Vallès),
- language change (Nerlich & Chamizo Domínguez),
- non-modularity of cognition and language and criticism of
the Chomskyan paradigm (Kravchenko's paper and the review of
his book, Paradis, Hamawand, Vallès),
- evolutionary biology (Kravchenko's paper and the review of
- applied CL (Caballero, Porto Requejo, Sánchez García,
Lakoff's interview), amongst others.
In the next section, the contributions are summarized. The
summaries are intentionally rather detailed, since this review
is, amongst others, meant to be a contribution to the
diffusion of CL in broader circles within the linguistics
community. In the discussion section, I discuss some of the
merits and shortcomings of the edition as a whole, and of the
papers in particular. As to the late, the discussion is
centered on the issue of methodologies, going beyond the scope
of the volume under review since it is intended to summarize
the main tenets of the CL enterprise and of the methodological
desiderata which should be met in order to improve its
descriptive and explanatory powers, which by the way are not
small. Two other relevant points are concerned with cross-
disciplinarity and application of CL, which are nevertheless
not fully discussed here.
I would like to thank the following people, including some of
the authors of the volume under review, for stimulating
discussions, fascinating papers, and valuable clarifications,
by means of personal communications and/or contributions to
the Cogling discussion list (I am particularly obliged to the
people who responded to my query (Jul 04 2004) to the Cogling
concerning the methodologies used in CL): Dmitry G.
Bogushevich, Per Aage Brandt, Giancarlo Buoiano, Rosario
Caballero, Israel ''izzy'' Cohen, Phillip Elliott, Dirk
Geeraerts, Zeki Hamawand, Stefan Gries, Tarik Hadzibeganovic,
John Hewson, Joe Hilferty, Priscilla Hill, Anders Hougaard,
Iraide Ibarretxe-Antuñano, Rembrandt Klopper, Manfred Krifka,
George Lakoff, Sydney Lamb, Adam E. Leeds, Robert ''Beau'' Link,
Ana Ortigosa Pastor, Ma. Dolores Porto Requejo, Oren Sadeh-
Leicht, Chris Sinha, and Robin Turner. I am sorry if I have
forgotten anyone. And I am sorry that, for limitations of time
and space, I could not include all their contributions and
references in this review.
SUMMARY OF CONTENTS
1. ''Towards a corpus-based identification of prototypical
instances of constructions'' by Stefan Th. Gries (Soenderborg,
Denmark) (pp. 1-27)
''The scope of [this] paper is [...] mainly methodological in
nature'', as Gries announces (p. 5). Taking as example the
analysis of the so-called 'dative alternation' constructions
in English from the perspective of the prototype theory of
categorization (e.g. Rosch 1973), Gries presents a method
which includes three phases:
(a) The analyst prepares an inventory of semantic, formal, and
pragmatic features which, according to a large number of
studies, determine the speaker's choice of construction.
(b) Then she analyzes a relatively representative corpus,
quantifying her findings so that she can identify the features
with high cue validity for prototypical instances of the
constructions involved, which enables predictions about the
speakers' choices of construction.
(c) Finally, she backs up her findings by means of natives'
In the case of the study in question, these phases looked like
(a) Gries' inventory of features which, according to other
studies dealing with the so-called 'dative alternation',
determine the speaker's choice of construction (ditransitive
or prepositional construction--DK and PK, respectively),
includes features such as the ''process described by the
utterance'', length and (pronominal or lexical) kind of NPs
realizing referents, animacy or discourse newness of
referents, amongst others (pp. 8f.).
(b) From files of the British National Corpus (BNC) Gries
extracted then 60 cases of DK and 57 cases of PK; by means of
a linear discriminant analysis (LDA) he could identify the
features with high cue validity for DK and PK, which enabled
him to characterize prototypical instances of each of both
constructions; the LDA has a high predictive power (88.9%)
(c) Finally, to back up his claims he conducted a
questionnaire experiment with English native speakers (p.
2. ''Entering in Spanish: Conceptual and semantic properties of
entrar en/a'' by Iraide Ibarretxe-Antuñano (then Bilbao, now
Saragosa, Spain) (pp. 29-58)
Also this paper deals with alternative constructions, namely
with the Spanish constructions 'entrar en' ('enter in') (in
these review called 'locative construction' or LK) and 'entrar
a' (in some cases + dative pronoun) ('enter to') (in these
review called 'allative construction' or AK) used to express
In the first part of the paper, Ibarretxe-Antuñano
characterizes entering events by means of the ''semantic
primitive of motion'' (pp. 29f., 43) and ''image schemas'' (e.g.
Johnson 1987, Lakoff 1987), viz. ''Source-Path-Goal (SPG)'' and
''Boundary (BND)'' (pp. 35f.). The choice of construction is
explained in terms of ''force dynamics'' (Talmy 1988),
''profiling of events'' (Langacker e.g. 1987) conveyed by the
prepositions 'a' (''dynamic'') and 'en' (''static''), and the
''polysemous character of the verb 'entrar''' conveying a
''dynamic'' ('to enter into') and a ''static sense'' ('to fit')
Ibarretxe-Antuñano concludes the first part by stating that,
while the ''primitive of motion'' is not irrelevant, the ''key
factors'' are ''the nature of the boundary crossing itself and
the force dynamic relation between the two entities'' involved
(p. 43). Thus, a LK can, for instance, convey a ''neutral force
dynamic relation'' between trajector and landmark, whereas an
AK can convey a ''negative force dynamic relation'' between
trajector and landmark which have the roles of agonist and
antagonist, respectively (pp. 39f.).
In the second part, Ibarretxe-Antuñano analyzes situations in
which she explains the choice of construction in terms of
metonymy, deixis, and scope. Thus, the LK can, for instance,
convey a concrete entering event (into a building), whereas
the AK can convey a metonymical understanding of the event
(metonymy ''ACTIVITY FOR PLACE'') (p. 47).
3. ''The construal of atemporalisation in complement clauses in
English'' by Zeki Hamawand (Hamburg, Germany) (pp. 59-85)
Starting from notions as ''temporal profile'', ''sequential
scanning'', and ''summary scanning'' amongst others, used to
characterize ''atemporalisers'' (Langacker e.g. 1987, 1991,
2000) (pp. 59ff.), Hamawand investigates the
conceptualizations of following atemporalizing
- ''the zero complementiser'': e.g. 'She made him clean the floor';
-'to': e.g. 'He decides to take early retirement';
- 'for-to': e.g. 'I want for Anna to meet the deadline'; and
- '-ing': e.g. ''Kate enjoys dancing the tango'.
After a critical review on the status of complementisers in
some other frameworks (pp. 62-64), Hamawand presents the
cognitive account which ascribes complementizers ''semantic
values''(p. 64), such as (1) ''conceptual distance'' (e.g. Haiman
1983) between main and complement events (pp. 66ff.); (2)
''temporal reference'' relating ''the time of a complement clause
to the time of the utterance'' (p. 61, pp. 68f.); as well as
(3) a polysemous character leading to conceptual networks
based on ''prototypicality and schemacity'' (Langacker 1987)
Thus, the example of ''zero complementiser'' above is explained
in terms of simultaneity of both events (pp.71ff.), whereas
the 'to' construction expresses posteriority of the complement
event, which is its ''schematic meaning'' that can have the
''extend meaning[s]'' of ''subsequent potentiality'' (e.g. 'They
hoped to climb the Mount Everest') and ''subsequent
actualization'' (e.g. 'She forced him to reconsider his
position') (pp. 73ff.). The complementizer 'for-to' has the
prototypical sense of ''subsequent potentiality'' as in the
example above, and the ''convoluted'' sense of ''coincident
actuality'' (e.g. 'It is wonderful for her to be in high
spirits') (p. 78).
In a sense, also this paper deals with alternative
constructions and their conceptualizations, since Hamawand is
concerned with both the speaker's ''conceptual flexibility to
construe a complex scene either temporally or atemporally'' (p.
61) ... see the definition of ''construal'' as the ''ability of the
speaker to conceive and express a situation in alternate ways''
(p. 59, abstract) -, and his choice of complementizer
depending upon semantic features of complementizers and
predicates (p. 65).
4. ''Talking about space: Image metaphor in architectural
discourse'' by Rosario Caballero (Spain) (pp. 87-105)
Caballero analyzes the occurrence of image metaphors in 95
building reviews selected from six magazines specialized in
architectural design. For this genre-based and corpus-driven
study, she developed a procedure in which the metaphors are
identified and analyzed semantic-conceptually according to the
respective underlying ''metaphorical mapping'' and
lexicogrammatically according to their recurrence in the
rhetorical structure. The aims of the study are ''to show (a)
the importance of image metaphors against commonly held views
on them in cognitive approaches and (b) point to the
difficulties of classifying metaphors into conceptual or image
types without any other consideration to their discourse
instantiation and function'' (the author in personal
Caballero found that the analyzed metaphors convey mainly
knowledge ''associations about the target based upon the
source'', such as FORM IS MOTION (e.g. 'meander'); others are
related to visually perceptible aspects (e.g. 'skeleton' or
''membrane'') (p. 92). The instances inform mainly nouns in
distinct ''types of pre-modification patterns'' (e.g.
'supermarket box' and 'pinwheel plan') (pp. 92f.). Verbs -
conveying a ''complex metaphorical transfer'' according to
conceptual metaphors such as BUILDINGS ARE ANIMATE BEINGS
(e.g. '[...] conservatories [...] which clamber up the crags [...]')
(p. 95) - and adjectives ... whose interpretation is mostly
contextually determined (e.g. 'reptilian' for the color
'green' (p. 96) - are also very productive (p. 93).
While the primary function of image metaphors in the building
reviews is informational, they also be used for evaluative
purposes. The informational role is a consequence of the
descriptive character of the genre. While metaphors can be
used referentially, replacing then the target term, they
''primarily work as extra-specification device'', with co-
occurrence of target and source terms. ''Image metaphor
clusters'' are used mainly to indicate different
perspectivizations of the buildings being described (''spatial''
or ''perspectival deixis''). (pp. 97ff.).
5. ''Temporal deictic adverbs: A constructionist approach'' by
Ana Ortigosa Pastor (La Rioja, Spain) (pp. 107-118)
Following ''the main tenets [sic] of Construction Grammar,
namely that each grammatical construction should specify
semantic, pragmatic and syntactic information'', Ortigosa
Pastor argues that a correlation of some notions posited in
Construction Grammar (CG) (Fillmore 2001), Role and Reference
Grammar (RRG) (Van Valin & LaPolla 1997), Functional Grammar
(FG) (Dik 1989/1997), and Lambrecht (1994) can be fruitful for
''an accurate description'' of the temporal deictic adverbs
'yesterday', 'today', and 'tomorrow' as ''constructional
templates'' (pp. 107f.).
Ortigosa Pastor expands (i) Fillmore's lexical-semantic
analysis of temporal adverbs (by means of a ''vector
construction'' with a ''temporal TARGET'', a ''temporal LANDMARK'',
and, for deictic adverbs, an ''anchoring landmark'', pp. 108-
111) by discussing (ii) the RRG's syntactic approach
(''temporal adjuncts are [...] modifiers of the core (and
sometimes of the clause - e.g. Yesterday, he went to the store
and bought some cheese), although similarly to operators they
take part in the operator projection'', the author in personal
communication); and (iii) the pragmatic perspective proposed
by Lambrecht and Dik : Ortigosa Pastor partially disagrees
''with Lambrecht's (1994) view, arguing that they often may be
part of the focus information (e.g. ''Secretary Pena is also on
his way out there. He was in Birmingham today'); in other
cases, they ''are the main focus of an utterance'', and ...
following ''Dik's (1989) typology of focus'' - convey ''some
contrastive information'' (e.g. 'We will start briefings for
you all perhaps as early as tomorrow, but definitely by
Wednesday') (pp. 113-116).
6. ''Del significado de la palabra a la interpretación del
texto: ¿Qué es la magia?'' by Ma. Dolores Porto Requejo
(Madrid, Spain) (pp. 119-135)
(Note: I quote passages from this Spanish text translating
them into English)
Porto Requejo examines the metaphors that constitute the
conceptualization of MAGIC in the Fantasy novel ''Forging the
Darksword'' (Weis & Hickman 1988) which presents a world,
'Thimhallan', whose inhabitants are born with magic powers ...
referred to in the novel as ''Magic'', ''Life'', or ''Life force'' -
, using them in their daily lives. Some are born without them,
being forced to make use of the ''Ninth Mystery'', ''Death'',
''Dark Art'', or ''Technology''', i.e., their own hands and hand-
Relying on Gibbs' (1998) notion that a concept is not a
monolithic entity, but a dynamic one which can be approached
from several perspectives, Porto Requejo argues that, for the
investigation of the concept of MAGIC in the mentioned novel,
its distinct ''metaphorical projections'' must be considered (p.
121), since they make up a ''complex network of metaphors,
absolutely structured and hierarchically organized that
provides us with a global mental image of MAGIC'' (p. 119,
abstract, p. 122, pp. 130ff.). This ''global mental image'' is
based on a ''megametaphor'' (Werth 1994) or ''master metaphor''
(Kövecses 2000), as the ''central metaphorical projection'' from
which the others are derived (p. 119, pp. 132ff.).
In the novel under study, the concepts of MAGIC and TECHNOLOGY
are presented in an opposition by means of metaphors such
MAGIC IS LIFE/NATURAL ORDER and TECHNOLOGY IS DEATH/CHAOS (pp.
126-129). This opposition result in a '''positive-negative'
evaluation system'' very common in many concepts in real life
(Kövecses 2000) and also in the epic fantasy genre. For the
novel under study, these specific metaphors may be reduced to
two generic metaphors: MAGIC IS THE GOOD and TECHNOLOGY IS THE
EVIL (pp. 129f.). The underlying ''megametaphor'' reads ORDER IS
THE GOOD, CHAOS IS THE EVIL. Relying on Lakoff's & Turner's
(1989) notion of ''persuasive force'' of ''poetic metaphors'' as
extensions of ''daily metaphors'' (p. 121), Porto Requejo argues
that the mentioned ''megametaphor'' is that which confers
credibility on the novel since it is common for both the real
and the fictional world (pp. 133f.).
7. ''Lexical creativity and the organization of the lexicon'' by
Teresa Vallès (Spain) (pp. 137-160)
Partially on the basis of corpus data, Vallès investigates
Catalan neologisms, combining Bybee's (1988) ''conception of
morphology as a _lexical organization_'' (p. 155, original
emphasis) and a reformulated version of van Marle's (1985)
theory on morphological paradigmatic relations framing the
lexicon as start point for the study of lexical creativity
(pp. 137, 141, 150, 155), as well as adapting Bybee's (2001)
network model to the study of lexical morphology processes of
derivation and analogy (pp. 137, 138, 141, 143).
Discussing critically generative ''models of possible words''
(pp. 137f., 139-141, 150, 155), Vallès presents ''a usage-based
model of actual words'' (pp. 138f., 140, 141f.), aiming to
demonstrate that the study of neologisms provides insights
into the morphological organization of the lexicon, and that
the study of the lexicon's morphological organization
''constitutes an enriching approach to lexical creativity'' (pp.
137, 138, 141, 145ff., 146, 150, 153, 154). The interrelation
between lexicon and neology is explained in terms of networks
based on paradigmatic relations among words sharing morphemes,
''articulated over two main axes: relations among words with
common affixes (derivational categories) and relations among
words sharing the same stem (word families)'' (p. 141).
Contrarily to the common view that considers affixes (and,
consequently, derivation) ''as the key to lexical creativity''
since they are more productive than stems, Vallès argues that
neologisms (e.g. _autoeditor_ 'self-publisher') comprise both
stems and affixes, and are therefore the result of the
productivity of both (p. 151).
The paradigmatic relations are ''expressed by means of
morphological rules or _patterns emerging from the intrinsic
organization of the lexicon_'' (p. 146, original emphasis). For
instance, relations between derivational categories by means
of affix attachment are based on ''low-level patterns''
(Langacker, e.g. 1987) specifying, for instance, that the
adjectival suffix _-ble_ always appears after the basal
thematic vowel (e.g. _agradable_ 'pleasant') (pp. 146-150).
Vallès discusses the distinction traditionally made between
''rule-based derivational process'' (affix attachment) and
analogy (affix replacement, e.g. words with the semantically
related prefixes _macro-_ and _micro-_), arguing with Motsch
(1988) for a rather ''fuzzy boundary between derivation and
analogy'' (p. 139, 156).
8. ''Amor y metáfora conceptual: Aproximación a los sonetos 153
y 154 de Shakespeare desde la lingüística cognitiva'' by Manuel
Sánchez García (Cáceres, Spain) (pp. 161-177)
(Note: I quote passages from this Spanish paper translating
them into English)
Basing the investigation on (a) the distinction between
''conceptual metaphor'' and ''metaphorical expression'' (Lakoff
and Turner 1989) (pp.162-164); (b) the partial nature of
metaphorical projections with the concepts of ''highlighting''
and ''metaphorical utilization'' (Kövecses 2002) (pp. 164f.);
and (c) the assumption that Shakespeare wrote the mentioned
poems as poetic exercises (two distinct sonnets on the same
theme), Sánchez García aims at showing that in these poems
Shakespeare used distinct ''metaphorical expressions'' to
express a reduced number of ''conceptual metaphors'' (pp. 167,
The conceptual metaphors are LOVE IS FIRE, LOVE IS A SICKNESS,
LOVE IS HEAT, and LOVE IS SLAVERY, in which distinct aspects
of the target domain are highlighted in that some
characteristics of the source domains are metaphorically
utilized (p. 168). Shakespeare used, amongst others, following
''metaphorical expressions'': ''his [Cupid's] heart-inflaming
brand'', or ''against strange maladies a sovereign cure'' (pp.
While analyzing literary texts, Sánchez García emphasizes the
notion that metaphors are not simply a literary figure of
speech, but a pervasive tool used in human thought and present
in both plain and poetic discourse (Lakoff & Johnson 1980)
(pp. 162, 166, 172, 173). In the case of Shakespeare's
metaphors, these were/are common in plain discourse (p. 172).
Accordingly, the author ultimately aims at providing a
practical example of application of cognitive linguistic
theory of metaphor (p. 168, 172).
9. ''The ontology of signs as linguistic and non-linguistic
entities: A cognitive perspective'' by Alexander V. Kravchenko
(Irkutsk, Russia) (pp. 179-191)
Relying on Maturana (e.g. 1987) Kravchenko outlines his
''biocognitive philosophy of language'' (p. 187), presented in
Kravchenko (2001) (see section 13 below) within the framework
of ''autopoiesis as a new epistemology of the living'' which ''is
founded on a biological concept of cognition and language'' (p.
180). This approach provides ''an effective alternative'' to the
''traditional philosophical/linguistic analysis of semiotic
phenomena'' which is based on ''false epistemological
assumption[s]'', such as that linguistic and non-linguistic
''entities'' are ontologically distinct (pp. 179f., 182, 185,
188). According to Kravchenko, this view offers ''a genuinely
scientific (naturalist) angle and is a giant step toward
understanding consciousness and cognitive (mental) processes''
In autopoiesis, cognition ''serves an active organism in its
adaptation to its experiential world'' (p. 180). Communication
takes place by means of both coded and non-coded signs, as
gestures or smells (p. 182), it ''is an operational cognitive
domain of interactions in which the interacting organisms
cause orientational modifications in each other's behaviors''
(p. 183), whereas language is ''just one among many possible
domains of orientational interactions'' (p. 183). Semiotic
relationships are based on ''mutual causality'', i.e., ''_a word
may function as a sign of some entity, an entity may function
as a sign of the word_'' (p. 184, original emphasis). Both
linguistic and non-linguistic ''entities'' are (a) equally
''empirical objects'', ''natural constituents'' which, ''on the
grounds of their ostensibility/perceptibility'', belong to the
''environmental niche'' or ''interactional domain'' of the Homo
sapiens ''as living organisms'' (b) ''contribute to the formation
of a single concept on equal epistemic grounds'' (c) resulting
in mental representations (pp. 184, 185), conceived of as
''relative neuronal activities characterizing the state of an
organism's nervous system as a structure-determined system''
(p. 181, 185).
10. ''The use of 'literally': Vice or virtue?'' by Brigitte
Nerlich (Nottingham, UK) and Pedro J. Chamizo Domínguez
(Málaga, Spain) (pp. 193-206)
Starting from the polysemous and/or ambiguous use of the word
'literally' in present-day English as both 'unfigurative' and
'figurative' (pp. 196, 198), as in the example ''It's literally
freezing out there'' (p. 194), Nerlich & Chamizo Domínguez
analyze the use of the word 'literally' from three
perspectives - diachronic, normative, and synchronic -, basing
the investigation on the ''Oxford English Dictionary'', public
texts as web articles, the normative ''New Fowler Modern
English Usage'' (1968), and texts from the ''Bank of English''.
Normatively, the figurative sense is often dismissed in public
debates in terms of ''destruction'' of the English language. The
authors reject such attitudes by arguing ''that the English
language [not only] is tougher than purists [...] give it credit
for'', but ''has even profited from'' ''such semantic onslaughts''
(pp. 193f., 195, 198ff.)
Diachronically, the unfigurative sense is the oldest (emerged
in the early 16th century) and the ''prototypical'' meaning,
whereas the figurative sense emerged only in the late 19th in
the USA and became ever since very common (pp. 195, 198). What
some call ''deterioration'' is rather ''the natural result of
various well-known mechanisms of semantic and pragmatic
change, such as subjectification, bleaching, and a gradual
strengthening of the rhetorical stance in the uses of
'literally''' (p. 198).
Synchronically, 'literally' can modify the meaning of the
expression within its scope in three main ways: (a) 'in the
literal sense' as synonymous to 'faithfully' or 'precisely' or
'really'; (b) ''a gradual shift in meaning'' when used for
''rhetorical purposes'' with the ''intersubjective function'' of
''emphasiz[ing] the literal meaning of an almost 'dead'
metaphor'', i.e., of an idiom; and (c) a ''metaphorically
hyperbolical'' meaning, with a ''complete shift in meaning'',
being synonymous with 'virtually' (pp. 194, 195f., 201ff.).
11. ''Is the notion of 'linguistic competence' relevant in
Cognitive Linguistics?'' by Carita Paradis (Lund) (pp. 207-231)
In this paper, ''meant to be a philosophical contribution to
linguistic theorizing'', Paradis argues ''that the question of
the relevance of linguistic competence [as posited by Chomsky]
is a non-question'' from the Cognitive Linguistic (CL)
perspective, since it is a ''theoretical construct'' based on
the ''empirically arbitrary'' assumption about the modularity of
cognition and language, whereas in CL ''[l]inguistic and non-
linguistic (encyclopaedic) knowledge'' lie on a continuum and
semantic and conceptual aspects are crucial (pp. 207f.,
Paradis reviews critically the generative ''narrow view of
language'', qualifying its assumptions and methodology as
''counter-productive for the development of the theory'' since
they reduce the aimed descriptive and explanatory adequacy
(pp. 209-211, 225ff.).
In opposition, in the CL approach human cognition is conceived
of as a network, and [l]anguage as an integral part of'' it;
the study of language is therefore connected to psychology,
cognitive science, and neurology (p. 212). Perceptions and
bodily experience are crucial for the formation of conceptual
structures and, consequently, for both language acquisition
and language use, and the linguistic input is ''extensive and
highly redundant'' (p. 213, 214).
Since there is no divide between linguistic and encyclopaedic
meanings/knowledge (p. 212, 215), natural languages exhibit a
great ''combinatorial complexity'', as demonstrated in studies
on polysemy ''using real data'' (pp. 219ff.). In this respect,
Paradis adheres to Pustejovsky's (1998b) and Langacker's
(1999) ''weak polymorphism'' according to which polysemy is both
lexically and pragmatically determined, but ''[w]ords in
context are prone to evoke more meaning specifications'' (p.
221). Relying on Murphy (2000) for the analysis of the ''nature
of lexical knowledge'', Paradis posits a ''two-level model of
lexical knowledge'' which includes for every lexical item: (i)
a ''conceptual knowledge _of/about_ the words'', and (ii) a
''conceptual knowledge _of/about_ the world'', in which
linguistic and encyclopaedic knowledge are inseparable (pp.
222-225), original emphasis).
Paradis concludes that competence in CL includes resources as
linguistic and encyclopaedic knowledge, as well as memory,
intentionality, and skills of abstraction (p. 226).
12. ''Interview: Language and cognition: George Lakoff on some
internal and external complexities'' by Jesús Sánchez (Córdoba,
Spain) (pp. 233-267)
In this interview from 2000/2001, Lakoff addresses several
issues crucial for the CL framework, as Neural Theory of
Language (NTL), Construction Grammar (CG), metaphor theory,
language acquisition, embodiment, amongst others. I chose to
collect some statements about the NTL since it is the thread
of the interview.
One of ''the current challenges'' for CL is the necessity of ''an
integrated theory''; NTL is conceived of as ''to create such an
integrated theory'' (pp. 233, 260, 263). The central question
in NTL is: ''How is it possible for a physical structure, like
the brain, which has just neurons that are connected, and fire
and work by chemistry ... how can chemistry produce ideas and
languages''? (p. 239). ''[...] we can model, via computational
means, how those electric chemical connections work'' and ''we
also, through the study of the embodiment of mind, have been
able to get some ideas about what parts of the brain are
computing ... are characterizing certain concepts'' (pp. 239,
262). ''Regier, in his [...] model of image schemas, hypothesized
[...] that [...] parts of the visual system [...] can compute image
schemas [...]. It's a theory that explains how perception and
reason can be linked'' (pp. 239f., 245). ''[...] we were able to
[...] ground our cognitive linguistics in these computational
neural models, which are models of the actual chemical system
and ... as a result, the neural computation gives us a bridge
between the linguistics and the chemistry'' (p. 240). ''But [...]
it is not the case that the brain is a general computational
mechanism. [...] The brain puts constraints on what concepts can
be [and w]hat connections there can be [...]. [T]he brain puts
constraints on [...] image schemas in general, and [...] on the
possibility of metaphor and the learning of metaphors via the
neural learning mechanisms, namely recruitment'' ... this is the
''theory of neural learning'' (pp. 240, 245). ''[W]e are
discovering that the properties of neural systems are also
properties of linguistic systems'' (pp. 240f.).
13. BOOK REVIEWS
I will not, of course, review reviews, but I will give at
least a short account on what the books reviewed are about.
Review of Kravchenko (2001) by Victor I. Shakovsky (Volgograd,
Russia) (pp. 269-276)
This book discusses several issues in semiotics, linguistics,
and linguist semiotics from the perspective of Maturana-
Varela's (1980) ''autopoietic theory'', claiming ''that the
problem of cognition in philosophy cannot be resolved without
the resolution of the problem of linguistic meaning'' (see
summary of Kravchenko's paper in section 9 above).
Review of Ruiz de Mendoza Ibáñez et al. (2002) by Klaus-Uwe
Panther (Hamburg, Germany) (pp. 276-288)
This book provides an evaluative review of the research on
metonymy done up to date, develops a ''conceptual apparatus''
(in which metaphor and metonymy ''form a conceptual
continuum''), and applies this apparatus to grammar.
The Methodological Question
The CL approach is essentially a usage-based approach, which
means that it refuses some central axioms of the generative
paradigm, such as poverty of stimulus, innate Universal
Grammar (UG), modularity of human cognition and of language
knowledge, as explicitly or implicitly, and by all means
correctly, shown in the papers by Kravchenko and, above all,
Paradis. Paradis in her highly stimulating paper shows how
central and how ''counter-productive for the development of the
theory'' the methodological separation of linguistic and
encyclopaedic knowledge dictated by theoretical axioms is,
since it deprives the theory of its descriptive and
explanatory power (p. 211). Moreover, Paradis points out that
this separation is ''empirically arbitrary'' (pp. 207f., 225f.),
with which she means that the contextualized study of language
provides pieces of evidence for both the network nature of
human cognition and the assumption about linguistic forms
being the result of semantic, conceptual, and pragmatic
functions or structures, and not of (innate) algebraic rules.
The Chomskyan notion of ''linguistic competence'' has, as
properly demonstrated by Paradis, no utility in the CL
The implicit contention is that CL reinstates the social and
psychological dimensions of language which the
decontextualized study of the modular generative grammar had
disregarded. This is the reason why Geeraerts (2003)
distinguish ''decontextualizing and recontextualizing
tendencies in 20th century linguistics and literary theory'':
because the Chomskyan notion of competence ignores lectal
variations. This is also the reason why CL is assigned by
Harré's & Gillett's (1994) and Sinha (to appear) to what they
call ''second cognitive revolution'' and ''second generation
cognitive science'', respectively: because the notion of
modularity of human cognition has 'depsychologized' the
cognitive sciences. (Incidentally, I warmly recommend the
reading of Sinha's and Geeraert's papers because of their high
epistemological value and because they discuss the basic
tenets of CL from a historical perspective.)
The logical consequence of such a view would be a methodology
of contextualization in which the native speaker abstraction
and the introspective/intuitive method are dismissed. And
indeed, Paradis argues that, for instance, meanings can be
''generalized by abstracting away from contextual variants'',
but that ''[c]ontextual variants are more fundamental'' (p.
222), since ''[w]ords in context are prone to evoke more
meaning specifications'' (p. 221). That is the reason why she
defends the use of ''real data'', of contextualized data in
context, in the analysis of, for instance, polysemy (p. 219).
But the question is: ''What are real data? What are
contextualized data?'' Paradis (p. 219) compares the
decontextualized and the contextualized analysis of the uses
of the adjective 'old'. The decontextualized meaning is
'having long duration of time'. The contextualized meanings or
interpretations of 'old' are, paraphrasing Paradis'
expressions: (i) 'long in use' or 'not new' (e.g. 'an old
car'); (ii) 'long-lasting' (e.g. 'an old friend'); and (iii)
'former' or 'ex-' (e.g. 'an old boyfriend'). At the first
sight, there is nothing to be said against this analysis,
excepting perhaps that the adjective 'old' in 'an old friend'
seems to be related rather to the friendship than to the
friend, but this is not the point here. The point is: ''What is
the source of these data?'' Since information concerning the
source of the data is not given, it seems reasonable to
suppose that these are Paradis' own data, i.e., _self-
generated_ data. But can self-generated data be considered as
''_real_ data''? Does contextualization mean that the analyst
devices some situations in which the analyzed constructions
can be used?
To my mind, there is to these questions only one answer
possible: ''No''. And the reason is that self-generated data are
too near to the ideal speaker abstraction and the
introspective methode. And this might be dangerous, since the
analyst is in such situations more prone to ignore lectal
variation, even if she concentrates her analysis on conceptual
structures, as Paradis in her paper, which apart from that is
really inspiring, does.
But the problem is that this sort of introspective work seems
to be pervasive in CL. George Lakoff, for instance, who is one
of the founders of CL and one of the most deserving cognitive
linguists, responded to my timid query (Jul 04 2004) to the
discussion list ''Cogling'' concerning the methodologies used in
CL in the following terms: ''Since you have have [sic] a better
idea of what _you_ mean than you can have of what other people
mean, your semantic introspection is more likely to be
accurate when you are working on your own ''corpus'' than when
when [sic] you are working with other people's utterances''
(Jul 04 2004, original emphasis).
What would be the alternative to self-generated or pseudo-
empirical data? The alternative is: _really_ real data, really
_empirical_ data, as corpus data. Lakoff himself works with
corpus data, in his analyses of political texts (e.g. Lakoff
1996). Nevertheless, in his message (Jul 04 2004) to
''Cogling'', he considers: ''Corpus linguistics can only provide
you with utterances (or written letter sequences or character
sequences or sign assemblages). To do cognitive linguistics
with corpus data, you need to interpret the data -- to give it
meaning. The meaning doesn't occur in the corpus data. [...]
[t]his [i.e. corpus linguistics] must be done with the
recognition of the special difficulties imposed by having to
assign meanings to other people's utterances.''
To my mind, the meaning _does_ occur in the corpus data, this
is the work the cognitive linguist has to do: to find the
meaning in the empirical data. And in my view, the
''difficulties imposed by having to assign meanings to other
people's utterances'' should not be used as justification to
work introspectively (in the sense of non-empirically). What
we need are methods ''to assign meanings to other people's
utterances'', and this is exactly the point where the volume
under review gives some answers.
First of all, the great majority of the papers analyzing
linguistic data rely on corpus data or otherwise authentic
data, and merit to be called 'empirical papers', whereas the
'pseudo-empirical' or 'semi-empirical' papers are only three:
Ibarretxe-Antuñano's, Hamawand's, and Paradis' papers.
Paradis' paper was discussed above and I would like to stress
again the high quality of her analyses despite her
introspectivity. Hamawand's analysis of atemporalizers in
English is also very compelling (even though in some cases
other interpretations would be possible), but he analyzes
self-generated sentences only, and the ''evidences'' for his
interpretations are searched in grammaticality tests devised
by him, but not tested with native speakers, relying thus only
in his own intuitions.
Ibarretxe-Antuñano's analyses of constructions used to express
entering events in Spanish are also persuasive. To her credit,
it should be said that the data ''has been tested with Spanish
speakers from Northern Spain'' and the author points out that
dialectal studies are necessary to confirm her hypotheses (p.
55, n. 26). But: (i) Her first examples are very artificial,
which the author herself admits but does use for expository
reasons (e.g. 'El cuadrado entra en el triángulo', 'The square
enters the triangle') (pp. 33f.). (ii) Some other examples
were extracted from the 'Diccionario de la Lengua Española'
published by the 'Real Academia de la Lengua Española' (pp.
43-45) about which one might assume it is a rather
conservative, normative work. (iii) Other examples, again, are
self-generated. (iv) The author often uses expressions such as
'the construction xyz _triggers_ a different interpretation or
conceptualization' (pp. 29, 50, 51, 52, 53). To my mind, this
is an expression which reverses a main tenet of functional-
cognitive approaches, namely that constructions are the result
of (intended) meanings ... the author's expressions give the
idea that the meanings are the result of the constructions.
Kravchenko's philosophical paper is decidedly introspective.
Consider his supposition: ''If we try to conjecture an
organism's domain of interaction (its niche) as some physical
space constituted by a set of perceptible objects, the
proportion of linguistic objects found in it may turn out to
be negligible small if not zero'' (p. 185). Even if this is a
highly plausible conjecture, without empirical basis it
remains speculation only. I do not know if there is work
dealing with, say, input in children's daily life not only
concerned with child-directed speech, but also with non-
linguistic stimuli. But I would check it before I posit such a
hypothesis. Or I would write some lines on the necessity of
empirical evidence, even if I could not present any. As for
his speculations on categorization of linguistic and non-
linguistic ''objects'' (''What happens in the course of this
human's perception of (cognitive interaction with) these
objects?'' p. 185), there is abundant empirical literature on
it (see e.g. Tomasello 2003 for a review), but Kravchenko does
not mention any reference.
The truly empirical papers are: Gries (British National
Corpus), Caballero (architectural magazines), Ortigosa Pastor
(British National Corpus, Corpus of Spoken Professional
American English), Porto Requejo (Weis' & Hickman's fantasy
novel ''Forging the Darksword''), Vallès (the IEC [Institute for
Catalan Studies] Contemporaneous Catalan Corpus), Sánchez
García (Shakespeare's sonnets 153 and 154), Nerlich & Chamizo
Domínguez (Bank of English, web). And out of all of them,
Gries' paper is the most exemplary because of his efforts to
develop a method that aims at objectivizing the analysis.
Before discussing Gries' technique, let us consider the
extreme introspective method. The analyst:
(a) uses self-generated and decontextualized data (or data
from normative sources),
(b) analyzes these data introspectively, and
(c) without confronting his conclusions with native speaker's
These are, to my view, urgent problems which must be solved by
means of methodologies as empirical as possible. The situation
in (b) might be, in a sense, unavoidable, since some degree of
introspection is always required -- as Geeraerts (1999) in his
fascinating 'platonic' dialogue on ''idealist and empiricist
tendencies in cognitive semantics'' points out: in scientifical
work, regardless of whether the data are collected more
empirically or more introspectively, the interpretative aspect
is in the data analysis often, if not always, present and
decisive, and this hermeneutic job is not seldom strongly
influenced by the respective theoretical assumptions --,
nevertheless quantifying methods could reduce the degree of
introspection. And the situations in (a) and (c) could be
changed: (a) by means of naturalistic and/or experimental
data, and (c) by means of new methods.
In this context, Gries' study is really refreshing, because he
suggests some possible solutions to the three problems listed
above, even, partially, for (b):
(a) he uses naturalistic data and a large number of studies in
order to prepare an inventory of semantic, formal, and
pragmatic features determining the speaker's choice of the
constructions to be analyzed;
(b) he developed a quantitative method with a high predictive
power (this quantitative aspect mentioned in can also be found
in the work of Dirk Geeraerts' research group in Leuven who
also defends a more empirical approach; see e.g. Geeraerts 1999).
(c) he confronts his conclusions with native speaker's
There are some problems in Gries' method: (i) it can be
dismissed as ''number-crunching'' as Gries himself said in a
personal communication; (ii) at least in the paper reviewed on
dative alternation, the method and its presentation seem to
thrust the analysis itself into the background; (iii) Gries,
perhaps wisely, departs from the view that his
characterization of prototypical constructions should be
interpreted as the naïve native speaker's mental
representations or conceptualizations of such constructions,
and the ideal case would be that the linguist's
interpretations could be confronted with naïve native
speaker's judgments also in terms of mental representations.
However, Gries' work can be considered as highly significant
in the field of CL. His paper is, to my mind, also the most
relevant of the volume under review, followed by Paradis'
paper due to the methodological and the philosophical effort
of these respective authors to sharpen the contours of the CL
The ARCL -- Edition and Aims
First of all, I would like to make some remarks on the edition
as a whole. It should be said that, to my mind, the issue is
not sufficiently accurately edited. Firstly, the quantity of
misprints and 'ungrammaticalities' is high, reflecting perhaps
imprecisions of the original manuscripts which remained
unnoticed by the editors, Shakovsky's review of Kravchenko
(2001) being here a case in point (some 15 such inaccuracies
in 6 pages). Also, the presentation of some papers partially
lacks coherence, simplicity and precision; thus in some texts,
the authors fails to provide references (e.g. in Caballero's
paper, p. 91, it reads: ''Given the assumption in genre
research that [...]'' without to provide references); in Ortigosa
Pastor's paper, it is difficult to distinguish her own
contentions from the ones posited by the researchers she
cites; in this same paper, p. 113, the author discusses an
example ''(2f)'' that is not available; several passages in
Kravchenko's paper are somewhat obscure and perhaps also
contradictory). Secondly, the structure of the papers does not
always correspond to the ARCL guidelines. Thus, although
''[c]ontributions should be in English'', there are 2 articles
in Spanish; several papers are not ''accompanied by a
biographical note'' (e.g. Caballero, Vallès); in some articles,
the examples are not ''set apart from the main body of the
text'' (e.g. Hamawand); in some texts the notes are not ''kept
to a minimum'', Ibarretxe-Antuñano's paper being here a case in
point (33 notes); in some articles some central keywords are
missing (e.g. 'categorization' in Gries' paper), in others
there are no keywords (e.g. Ibarretxe-Antuñano). Thirdly, the
position of the notes at the end of the main text is
unpractical and some figures are not adequately placed (e.g.
Ibarretxe-Antuñano, Hamawand). Finally, the articles could
have been accompanied by the date of reception/production, and
the volume could have been completed by an index of names and
Apart from these rather 'microstructural' deficiencies, the
(macro)structure and presentation of the volume are very
satisfactory. Thus, the edition opens with considerations on
methodologies (Gries) and ends with a philosophical reflection
(Paradis) which are both central for CL, as it is discussed
below. To my mind, to judge by the quality and relevance of
the contributions of this first issue, the ARCL has by all
means the prospect of achieving the aim of ''establish[ing]
itself as an international forum for the publication of high-
quality original research on all areas of linguistic enquiry
from a cognitive perspective'', as it reads in the description
on the publisher's web site
and of achieving a position in the CL community similar to that
occupied by another high qualitative periodical, namely
''Cognitive Linguistics'' of the International Cognitive Linguistics
Association (ICLA) edited by Adele Goldberg, one of the leading
researchers in CL (see e.g. Goldberg 1995).
The ARCL, ''published under the auspices of the Spanish
Cognitive Linguistics Association'' (AELCO/SCOLA), which is
incidentally affiliated to the ICLA, can be seen as the result
of the increasing interest in CL in Spain. The editorial board
consists of Spanish scholars under the direction of Francisco
José Ruiz de Mendoza Ibáñez, one of the leading cognitive
linguists in Spain (see, for instance, his book (2002) on
metonymy reviewed in this first issue). In this respect, in
order to become ''an international forum'', it would be
desirable that the next issues of the ARCL renounces to the
predominance of Spanish authors which sets the tone of this
first issue (8 or 9 out of 15 authors).
Also, in order to achieve its second aim, namely to support
''[f]ruitful debate [...] with neighboring academic disciplines
as well as with other approaches to language study,
particularly functionally-oriented ones'', it would be good if
next issues contained more papers from cross-disciplinary
perspectives. For sure, Kravchenko's ''bio-cognitive philosophy
of language'' (p. 187) relies partially on evolutionary
biology, there are some papers in which CL is applied to
poetics (Porto Requejo, Sánchez García) and to the study of
genres (Caballero), and in Sánchez' interview Lakoff makes
some comments on the application of CL to cognitive therapy
(p. 266), and his own work on both NTL (see previous section)
and application of CL to ''social policy and to politics'' (pp.
264f.). It would be, nevertheless, very constructive to have
papers discussing CL tenets from the perspective of, for
instance, (i) developmental psychology (see, e.g., the work on
language acquisition and cognition done by Michael Tomasello
(e.g. 1997, 1999, 2003), Elena Lieven (e.g. Abbot-Smith et al.
2004), and associates in Leipzig; (ii) psychotherapy and
psychoanalysis (see, for instance, Barkfelt's 2003 study on
the therapeutic use of metaphors in the treatment of
depressions); (iii) brain research and evolutionary biology
(see, for instance, the paper by Crow 2000 ... admittedly based
on a generative conception of language - on the relations
between schizophrenia and language in the Homo sapiens); (iv)
neurolinguistics (see, for instance, Lamb 2003), to mention
some examples. Incidentally, this issue on cross-
disciplinarity and application of CL findings to both other
fields of scientifical work and spheres of human life is
highly relevant not only for the ARCL, but for the whole field
I think that the next issues of the ARCL has some desiderata
to meet of which some may be extended to the whole field of
CL. But all in all, this first issue can be seen as a
successful first step in the direction of the aimed
objectives: to ''establish itself as an international forum for
the publication of high-quality original research on all areas
of linguistic enquiry from a cognitive perspective'', and to
support ''[f]ruitful debate [...] with neighboring academic
disciplines as well as with other approaches to language
study, particularly functionally-oriented ones''. Moreover, it
is very recommendable for both readers with and without
knowledge in CL. For the late, the issue can even serve as a
sort of introduction into the field, since the majority of the
papers includes good explanations of the underlying notions
and common terminologies, even though parallel readings would
be by all means advisable.
The volume under review contains much more interesting and
relevant issues than the ones discussed in this review, e.g.
prototype theory, metaphor theory, alternate constructions, to
cite only some examples. I chose to concentrate the discussion
on the methodological issue since it is crucial for the field.
In this respect, I can conclude that, given the pervasiveness
of the extreme introspective method as described above, the
development of objectivizing methodologies is of central
relevance for the field of cognitive-functional linguistics,
in order to avoid that Paradis' criticisms to the generative
program concerning its restricted descriptive and explanatory
powers be applied to CL.
[Note: The references cited in the reviewed papers are not
included below; they can be provided on demand to the reviewer.]
Abbot-Smith, K., Lieven, E., & Tomasello, M. (2004): Training
2;6-year-olds to produce the transitive construction: The role
of frequency, semantic similarity and shared syntactic
distribution. Developmental Science, 7,1, 48 - 55.
Barkfelt, Judith (2003): Bilder (aus) der Depression.
Metaphorische Episoden über depressive Episoden: Szenarien des
Depressionserlebens. Konstanz: Hartung-Gorre.
Crow, T.J. (2000): Schizophrenia as the price that Homo
sapiens pays for language: A resolution of the central paradox
in the origin of the species. In: Brain Research Reviews, 31,
Geeraerts, Dirk (1999): Idealist and empiricist tendencies in
cognitive semantics. In: Janssen & Redeker (eds.), pp. 163-
Geeraerts, Dirk (2003): Decontextualizing and
recontextualizing tendencies in 20th century linguistics and
literary theory. In: Mengel, E., Schmid, H.-J. & Steppat, M.
(eds.): Anglistentag 2002 Bayreuth, pp. 369-379. Trier: Wiss.
Goldber, Adele (1995): Constructions: A Construction Grammar
Approach to Argument Structure. Chicago, IL: The Univ. of
Harré, R. & Gillett, G. (1994): The Discursive Mind. London:
Janssen, Theo & Redeker, Gisela (eds.) (1999): Cognitive
Linguistics: Foundations, Scope, and Methodology. Berlin: de
Lakoff, George (1996): Moral Politics: What Conservatives Know
That Liberals Don't.
Lamb, Sydney (2003): Neurolinguistics and General Linguistics:
The importance of the microscopic level. Logos and Language 4,
Kravchenko, Alexander (2001): Sign, Meaning, Knowledge: An
essay in the cognitive philosophy of language. Irkutsk,
Ruiz de Mendoza Ibáñez, Francisco José & Otal Campo, José Luis
(2002): Metonymy, Grammar, and Communication. Albolote: Ed.
Sinha, Chris (to appear): Cognitive linguistics, psychology
and cognitive science. In: Geeraerts, D. & Cuyckens, H.
(eds.): Handbook of Cognitive Linguistics.
Tomasello, M. & Call, J. (1997): Primate Cognition. Oxford
Tomasello, M. (1999). The Cultural Origins of Human Cognition.
Harvard University Press.
Tomasello, M. (2003): Constructing a Language: A Usage-Based
Theory of Language Acquisition. Cambridge, MA & London:
Harvard Univ. Press.
| ABOUT THE REVIEWER:
ABOUT THE REVIEWER
The reviewer is currently working on her M. A. thesis on the
acquisition of argument constructions in a bilingual child
within a usage-based framework. Her research interests include
first language acquisition, multilingualism, cognitive science,
developmental psychology, as well as history of linguistics.