| EDITOR: Amberber, Mengistu
TITLE: The Language of Memory in a Crosslinguistic Perspective
SERIES: Human Cognitive Processing 21
PUBLISHER: John Benjamins
Ludwig Fesenmeier, Romanisches Seminar, Ruhr-University Bochum
The purpose of the book under review is to explore ''the language of memory in a
cross-linguistic perspective'' (p. 1), i.e. the volume addresses the question of
''how different languages lexicalise the concept of memory'' (p. 5). The languages
taken into consideration are Polish, English, Dalabon, Russian, Chinese, German,
Korean, East Cree, and Amharic.
The book contains ten articles which are preceded by a section ''About the
contributors'' (pp. ix-xi) and an ''Introduction'' written by the editor (pp.
1-12); at the end one finds the ''Author index'' (p. 279), the ''Language index''
(p. 281), and the ''Subject index'' (pp. 283f).
The contributions represent revised versions of the respective papers presented
at the ''Workshop on the Semantics of Memory in a Crosslinguistic Perspective'',
held in November 2003 at the University of New South Wales (Sydney, Australia).
In the first article (pp. 13-39), Anna Wierzbicka answers the title question ''Is
'remember' a universal human concept?'' by saying that '''[m]emory' [...] is not
something that objectively exists'', but rather ''a construct, linked with the
current meaning of the English word MEMORY'' (p. 14, throughout this review, all
capital letters indicate italics in the original). Moreover, according to the
Author, memory would encode a quite complex concept which, thus, has to be
decomposed in simpler terms in order to both demonstrate its language-specific
semantic configuration and allow for crosslinguistic comparison. The analytical
tool proposed is the ''Natural Semantic Metalanguage'' (henceforth NSM), currently
consisting of around 60 semantic primitives (see p. 18), the most relevant here
being /think\ and /know\ [Editor's note: slashes are used in this review in
place of angled brackets]. After having briefly mentioned some of the
expressions for 'memory' and 'to remember' in different languages (Australian
Aboriginal languages, French, German, and Russian) and after a short overlook
over the semantic change of the English verb TO REMEMBER, Wierzbicka discusses
in some detail ''some Polish words related to 'memory''' (p. 29), concluding that
''most Polish words in this area do not match the meanings of English 'memory
words''' (p. 37).
John Sutton's paper ''Language, memory, and the concepts of memory'' (pp. 41-65)
comments on the NSM approach (more precisely, on Wierzbicka's contribution) on a
methodological level, stating first of all the many different kinds of inquiries
in the field of 'memory' and related concepts: there is a great amount of
research done especially in philosophy and cognitive and developmental
psychology. The Author then goes on to reject a key assumption which parallels
that made by Wierzbicka, concerning the very ''nature'' of memory: ''a number of
writers [within science studies and the history of psychology] have used
evidence of HISTORICAL variation in the constitution of 'memory' to argue that
memory is not a natural object or a natural kind, in just the same way that
Wierzbicka uses cross-linguistic evidence'' (p. 47). In Sutton's view, such
dichotomies as 'nature vs. discourse' or 'science vs. history' are simply
irrelevant when ''appreciat[ing] and utilis[ing] cross-linguistic,
cross-cultural, or historical evidence of substantial diversity'' (p. 48).
Furthermore, the Author addresses Wierzbicka's explanation of certain English
words, suggesting that they are ''in certain different respects too strong, in
going beyond the basic semantics of the English terms by building in too much
idiosyncratic metaphysical baggage'' (p. 53). He equally questions the ''deep
differences between Polish and English attitudes to the past'' (p. 60) proposed
by Wierzbicka on the basis of her analysis of Polish PAMIATKA (see pp. 32-35).
In conclusion, Sutton argues for ''a broader interdisciplinary enterprise of
coming to understand thinking about what happened before'' (p. 61).
In his contribution ''Standing up your mind. Remembering in Dalabon'' (pp. 67-95),
Nicholas Evans explores the possibilities of linguistic expression in the
cognitive domain in Dalabon, an Australian Aboriginal language, ''now spoken
fluently by fewer than a dozen people'' (p. 67), where ''expressions specifically
dedicated to remembering'' (p. 68) are completely missing: there is, rather, the
bound root BENG, generally covering ''the whole realm of mind'' (p. 76), which is
the morphological basis for verbs such as BENGDI and BENGKAN, denoting, among
others, the concept of 'to remember'. The Author proposes to treat these verbs
not as polysemous, but rather as monosemous, presuming ''a single meaning for
each, from which the various contextual readings emerge from interactions with
other grammatical elements in the verb (particularly the tense, aspect and mood
inflections) and from other aspects of context'' (p. 84); in the case of BENGDI,
this single meaning can be glossed as 'to have a thought or experience
temporarily in one's conscious mind' (see p. 85), while BENGKAN can be
paraphrased as '''[to] have continually in one's mind''' (p. 86). Finally, in a
quite intriguing ''Appendix'' section (pp. 91-94), Evans addresses the question
whether in Dalabon /think\ and /know\ can be treated as semantic primitives in
the NSM sense. His answer is negative, for according to the Author, this
language lacks verbs meaning PRECISELY 'think' and 'know' (see p. 92). He thus
makes a strong case for keeping the ''fundamental assumptions of the NSM school
of semantics'' strictly distinct: ''(a) each language can be used as its own
metalanguage [...] (b) the set of 'semantic primitives' yielded by (a) in each
language is isomorphic and directly intertranslatable'' (p. 91).
Anna A. Zalizniak draws attention to ''The conceptualisation of REMEMBERING and
FORGETTING in Russian'' (pp. 97-118), which according to another analysis,
affirmatively quoted by Zalizniak, follows the metaphor of possession/loss (see
p. 98). As for remembering, the Author argues that an opposition exists between
verbs denoting states (imperfective POMNIT'), processes (imperfective
VSPOMINAT'), and events (perfective ZAPOMNIT' and VSPOMNIT'); concerning
forgetting, Russian appears to use three quasi-metaphorical expressions: ''1)
[...] momentary dropping out of an object [...]; 2) a gradual disappearance of
'signs' of the life experiences 'written down' in the mind [...]; 3) the
covering with something like mist'' (p. 110). Discussing the possible semantic
decomposition of POMNIT', Zalizniak shows that at least for certain readings, a
semantic primitive /know\ has to be included, since in some cases POMNIT' can be
substituted by ZNAT'.
Cliff Goddard offers ''A 'lexicographic portrait' of FORGETTING'' (pp. 119-137;
italics in original), examining the subcategorization frames of English TO
FORGET with special interest to their possible semantic motivations. The
analysis is mainly corpus-based (COBUILD corpus) and uses the NSM technique for
the semantic description. The Author discusses in turn the three main clausal
complement types (TO FORGET TO/THAT/WH-complement), NP-complements, TO FORGET
ABOUT, TO FORGET associated with certain kinds of modifiers (''I'll never
forget'', ''I can't forget''), and some idiosyncratic uses (e.g. ''Forget it!'', TO
FORGET ONESELF). Goddard points out parallels between clausal and nominal
complement constructions and identifies a total of fifteen different, though
''interrelated lexicogrammatical constructions [of TO FORGET], each with a
specific meaning'' (p. 134). In his ''Concluding remarks'' (pp. 134f), he stresses
the necessity of a joint approach, where semantic analysis ''make[s] sense of
[...] both syntactic options and actual usage patterns'' (p. 135).
The article of Zhengdao Ye (pp. 139-180) is entitled '''Memorisation', learning,
and cultural cognition'', more specifically being devoted to ''The notion of BÈI
('auditory memorisation') in the written Chinese tradition''. The Author analyzes
in detail the Chinese expression BÈI, which is of central interest here, and
proposes a distinction between two different, yet metonymically related
readings: BÈI-1 focuses on the act of memorization itself, while BÈI-2 denotes
the result of BÈI-1. Furthermore, BÈI appears to be generally restricted to the
realm of verbal learning, while this is not the case with its semantic
near-equivalent JÌ, ''which is akin to 'taking a mental picture', using objects
that are not restricted to 'texts''' (p. 175). Ye applies the NSM approach in
order to present the meanings of these highly culture-specific concepts in such
a way that they become ''accessible and intelligible to people from other
language and cultural backgrounds'' (p. 148).
''A corpus-based analysis of German (SICH) ERINNERN'' (italics in original) is the
topic of the paper by Andrea C. Schalley and Sandra Kuhn, which, in contrast to
what its title suggests, is not based on usage or on syntactic structures (as is
Goddard's, see above), but aims ''to identify rigorously what is actually part of
the lexical semantics of the verb (SICH) ERINNERN'' (p. 182f; italics in
original); the examples discussed, however, are taken from the COSMAS II corpus.
The framework used is that of ''Unified Eventity Representation'', based on the
''Unified Modeling Language'' (see p. 182); the approach chosen is called
''object-oriented'' (p. 182): at the centre of this model lies ''the concept of an
object or entity, whose characteristics, relations to other entities, behaviour,
and interactions with other entities are modelled'' (p. 182). The Authors start
their analysis by discussing in some detail the study of Van valin & Wilkins
1993, and then propose ''prototypical components of REMEMBER [capitals in
original], where REMEMBER [capitals in original] is understood as the eventity
which comprises the 'core semantics' of (SICH) ERINNERN [italics in original]''
(p. 192). Applying this analysis to German (SICH) ERINNERN, Schalley and Kuhn
identify four different readings of the verb: an ''achievement ERINNERN'' and its
causative variant, an ERINNERN 'to be similar to' and an ERINNERN used to make
the addressee ''aware of 'what she should know''' (p. 197). The Authors conclude
that ''there is some non-neglectable semantic difference but that there is no
clear-cut reading difference'' (p. 205).
In '''Do you REMEMBER where you put the key?' The Korean model of REMEMBERING''
(pp. 209-233; italics in original) Kyung-Joo Yoon deals, on the basis of NSM,
with the ''translational counterparts of REMEMBER'' (p. 209; italics in original),
showing that none of the Korean expressions taken into consideration mean
exactly the same as the English verb TO REMEMBER. The items focused on are
SAYNGKAKNA- 'come to think, be reminded of', KIEKNA- 'memory comes', KIEKHA-
'remember, recall', and KIEKHAYNAY- 'manage to remember', which can be
considered ''basic in terms of frequency of use, complexity of meaning, and
morphological structure'' (p. 215). According to the Author, KIEKNA- and
KIEKHAYNAY- differ from KIEKNA- in that the former two express the speaker's
volition to remember, while the latter ''implies involuntary retrieval'' (p. 225),
KIEKHAYNAY- nevertheless focusing more strongly the deliberate character of the
retrieval than does KIEKHA-. Concerning the semantically very close expressions
KIEKNA- and SAYNGKAKNA-, Yoon suggests that the former ''entails a past time
reference intrinsically'' (p. 229), while in the latter case, such an
interpretation seems to be due to inferences drawn from the context.
The following article by Marie-Odile Junker deals with ''The language of memory
in East Cree'' (pp. 235-261), an Algonquian language, which appears not to have
''a specific word for 'memory', but rather a hyperonym that encompasses all
mental processes'' (p. 259); nevertheless, one finds several verbs which ''are
used to express concepts corresponding to the English words 'remember',
'forget', and 'remind''' (p. 241). Some of them have in common the root
CHISCHIS(I)-, e.g. CHISCHISÛ 'to remember something from the past', which allows
for three different constructions: with the conjunct preverbs CHECHÎ, E, and KÂ,
respectively. According to the Author, these constructions can be paralleled
with the English TO + infinitive construction, the accusative -ING construction,
and the THAT construction, and can thus be classified, in line with the approach
of Van valin & Wilkins 1993, as PSY-ACTION, DIRECT PERCEPTION, and COGNITION
COMPLEMENT (see p. 242f). Furthermore, Junker shows how in East Cree the concept
of 'memory' is also an important category for the description of certain
grammatical items: the felicitous use of a certain class of demonstrative
pronouns (called ''absentative demonstratives'', p. 251) requires both the speaker
to remember the entity referred to and the hearer to agree on such a (virtual)
''presence'' by remembering (see pp. 251-253); there are also special verb forms
which are used ''when the speaker knows and remembers what something or someone
looked like before, but now sees that it has changed'' (p. 254), but are
incompatible with sentences which do not presuppose such knowledge (see ex. 90
and 91 on p. 255, glossed as ''I do not know that/if it was white'').
The final paper by Mengistu Amberber on ''REMEMBER, REMIND, and FORGET in
Amharic'' (pp. 263-277) first examines three Amharic memory verbs, corresponding
roughly to the English REMEMBER, REMIND, and FORGET, respectively; the Author
then discusses some verbs which normally mean 'to think'/'to know', but
occasionally can also be used in the sense of 'to remember', underlining the
fact that such polysemy does not, however, expand to the domain of 'forgetting'
(see p. 275). Among the many interesting issues which emerge from the analysis,
the most intriguing one appears to be the grammaticalization of the distinction
between volitional memory (''one is actively trying to recall'', p. 266) and a
simple remembering act that happens to the experiencer. The methodological
framework appeals to NSM and Jackendoff's Conceptual Semantics.
The different contributions treat a wide range of phenomena, showing that a
great and impressive variety of language-specific solutions exist for the
challenge of how to say ''what has happened before'', as John Sutton has put it.
When comparing the results of the analyses of near-synonyms, languages turn out
to differ with regard to the distinctions they make by lexical and/or
lexico-grammatical means (e.g. the rather ''curious'' distinction between
''remembering an habitual event and remembering a particular event'' in East Cree,
p. 243); nevertheless, one also finds striking crosslinguistic parallels, e.g.
''the distinction between intended and 'mind-popping' retrieval'' (p. 225) in
Korean and that between volitional memory and the simple remembering act that
happens to the experiencer in Amharic.
A common methodological denominator of many of the contributions consists in
that they appeal to the decompositional NSM approach, basically accepting it as
a suitable framework for the analyses to be undertaken. Differences mostly
concern ''details'' such as the status of certain primitives (/think\ and /know\)
in the analysis of the memory words of a given language (see, e.g., p. 103 for
Russian and p. 190 for German); nevertheless, one also finds views which are
rather globally opposed to some foundational issues related to the NSM approach
and the study of 'memory' within such a framework (see Sutton's comment on
However, the volume under review not only brings together different languages,
but also different theoretical frameworks: besides the NSM approach, articles
frequently make reference to Van Valin & Wilkins 1993 (see, among others, the
articles by Schalley & Kuhn, Yoon and Junker), but one also finds more recent
frameworks such as Unified Eventity Representation, applied to German data by
Schalley and Kuhn.
Furthermore, some of the papers (Goddard, Junker, Amberber) can be seen as
continuing previous work done on the semantically very close field of 'think'
(see volume 14-2/3 of ''Cognitive Linguistics'', explicitly referring to p. 2)
thus allowing for a more comprehensive view on the ''broad field'' of cognitive verbs.
As far as the several contributions are concerned, seeing that for reasons of
space I cannot go into great detail here, I shall confine myself to some brief
Wierzbicka quotes the statement made by Umberto Eco that 'remembering' probably
is one of the ''elementary notions, common to everyone in the human race, that
CAN BE EXPRESSED in all languages'' (see p. 13; emphasis mine), but this is not
to say that Eco ''is convinced [...] that all languages must have a WORD for it''
(p. 20; emphasis mine): a point in case is Dalabon, which, according to Evans,
''offers a number of distinct ways of talking about remembering [...] but without
having any lexicalised verb for 'remember''' (p. 90).
Admittedly, ''linguistics has not contributed so fully to the interdisciplinary
study of diversity in ways of thinking about what happened before'' (p. 44), but
this is NOT due to the relative neglect of semantics, ''at least until recently'',
as Sutton suggests: semantics not only represented an important linguistic
subdiscipline for quite some time, but also, already as early as 1966, Kurt
Baldinger had examined ''The Concept 'remember' and its representations in
French'' as ''An Example of Conceptual Analysis'' (see Baldinger 1980, 160-205). It
is rather that cognitive linguistics/semantics is a relatively recent paradigm,
especially committed to the possible relations between cognition and language(s)
(note, nevertheless, that Baldinger's approach is structuralist in nature).
Consider the following two statements by Wierzbicka and Schalley & Kuhn,
respectively: ''['Memory'] is a construct, linked with the current meaning of the
English word MEMORY - a construct that many psychologists and cognitive
scientists tend to reify by treating it as something that 'exists' independently
of the English language'' (p. 14); ''memory is not seen as something that exists
objectively (Wierzbicka, this volume)'' (p. 187). While Wierzbicka rightly
insists on the language-specificity of WORD MEANINGS which may not be used
straightforwardly as a ''neutral measure'' when discussing mental phenomena or the
(extra-linguistic) concepts such WORDS refer to, Schalley & Kuhn seem to go a
step further in generalizing this view.
As Evans rightly states (see p. 90), the study of lexemes pertaining to the
cognitive domain is particularly difficult because of their non-ostensible
denotations. It appears thus more important to distinguish, both on the
conceptual and terminological level, between what is linguistic and what is
extra-linguistic. When reading the different contributions, one is confronted
with expressions such as ''word'', ''meaning'', and ''concept'', the latter sometimes
being clearly used as a synonym for 'word(-meaning)', sometimes clearly in the
sense of 'conceptual entity'; yet sometimes one may have one's doubts: ''[the
matter of current concern is the predicate 'be.in.mind' used by Van Valin &
Wilkins 1993] Yet, MIND is an English-specific concept and does not have exact
counterparts in, e.g. German or Korean'' (p. 188).
As far as the formal aspects of the volume are concerned, one notes
- some typographical errors (among others ''je me ra[p]pelle'' (p. 22), ''se
rappelLer'' (p. 26), ''pos[s]essing'' (p. 97), ''reZervation'' (p. 111), ''LoquoUr,
ergo sum?'' (p. 117), ''While this MAYBE due to'' (p. 270));
- probably ''remnants'' of earlier versions of the papers: ''Dalabon is a
polysynthetic VERB [language?] with a complex verb structure'' (p. 70), ''the
analysis OF THE proposed by Wierzbicka'' (p. 105);
- the quotation p. 67 taken from Auroux 1989, 14 (not 134) reads correctly ''Sans
mémoire et sans PROJET, il n'y a tout simplement PAS de savoir.'';
- sometimes incomplete bibliographical references (missing page references or
publishing places, see among others pp. 95, 117, 261);
- in Goddard's paper, from p. 123 onwards, two different font sizes are used;
the itemization of the constructions/explanations of TO FORGET ranges from ''A''
to ''P'', but it contains only 15 items since there is no explanation ''C'';
- the page numbers given for ''Latin'' in the ''Language index'' sometimes refer to
''Latin America'' (pp. 25, 36), those given for ''French'' sometimes refer to a
homonymous author (Robert M. French, p. 61) or to the WORD ''French'' in the
translations of some examples (pp. 115f, 186).
In conclusion, despite the problems discussed so far, this volume devoted to
''The Language of Memory in a Crosslinguistic Perspective'' is a book which has
long been sought and will serve the goals it is intended for most suitably.
Auroux, Sylvain. (1989) Introduction. In Auroux, Sylvain (ed.) _Histoire des
idées linguistiques. Vol. 1: La naissance des métalangages en Orient et en
Occident_. Liège/Bruxelles: Pierre Mardaga, 13-37.
Baldinger, Kurt. (1966) Sémantique et structure conceptuelle (le concept 'se
souvenir'). _Cahiers de lexicologie_ 8, 3-46.
Baldinger, Kurt. (1980) _Semantic theory_. Oxford: Basil Blackwell.
Van Valin, Robert D., Jr & Wilkins, David P. (1993) Predicting Semantic
Structure from Semantic Representations: REMEMBER in English and its Equivalents
in Mparntwe Arrernte. In Robert D. Van Valin, Jr. (ed.) _Advances in Role and
Reference Grammar_, Amsterdam/Philadelphia: John Benjamins (= _Current Issues in
Linguistic Theory_ 82), 499-534.
ABOUT THE REVIEWER
Ludwig Fesenmeier teaches Romance linguistics at the Department of Romance
Languages, Ruhr-University Bochum, and is currently working on his post-doctoral
thesis on lexical synonymy in the Romance languages.