Pu das Bär. An Examination of Gender and Translation in a Cognitive Linguistics Framework


0. Introduction

Gender in Cognitive Linguistics (hereafter referred to as CL) is a largely neglected aspect, as it does not easily fit the “image”-schemas of Cognitive Semantics. CL’s various theories, from Construction Grammar (CxG), with its semantic companion piece Scenes-and-Frames-Semantics (cf. Fillmore 1977)[i] to Cognitive Grammar[ii] (CG), have steered largely clear of gender as a linguistic category.

This neglect is, at least partly, due to the fact that CL is concerned with syntax rather than morphology and complex constructions of varying degrees of schematicity rather than their less complex companions. What’s more, gender markings are usually regarded as a clear case of pure phonology (cf. Taylor 2002:333f.) “contribut[ing] little to the symbolization of conceptual structure” (334).

Yet this view has lately been called into question as the ideological biases in inflectional classes have been revealed (cf. Nesset 2001[iii]). However, gender as a “semantically based” (Langacker 2002:304) grammatical phenomenon has already been analyzed as early as 1991 in CG, as explicated by Langacker (2002).

This will be taken as the starting point for this investigation in the possibilities and impossibilities of translating gender. Approaches by Lakoff (1987) and Tabakowska (1993b) will be used, but central to this paper will be the Cognitive Linguistic view of gender, as the “translation muddle” (Hyde 1993:3) might profit from this perspective. After laying down the foundations for the arguments we will briefly examine the linguistic category of gender and how the Sapir-Whorf-Hypothesis might be applicable to it. Only then will we introduce the text, which is to be used as corpus: The Complete Winnie-the-Pooh by A.A. Milne[iv] and compare it regarding the character ‘Owl’ with the German standard translation by the esteemed Harry Rowohlt.

How categories of language influence understanding, writing or translation, if there exists something akin to a ‘faithful’ translation or if, ultimately, the Italian proverb that ‘traduttore traditore’[v] is right, these questions will be answered, though tentatively, at the end of this paper.

1. Cognitive Grammar

1.1. The framework (Spanish)

Clearly, CL “is not [...] the same as cognitive grammar” (Langacker 2002:ix), but in this section we will not take CxG into account at all, as its description of gender is not as useful as the CG version. Owing to Langacker’s discussion of gender being mostly restricted to Spanish gender[vi], we will subsequently slightly adapt it in order to fit the German inflectional classes as well.

The first thing that has to be noted is that Spanish nouns have a very strict adherence to gender. Whereas in German gender seems to be largely arbitrary (not fully, cf. Köpcke and Zubin 1984), with nouns of different grammatical gender like Feder (female) and Kater (male) providing no clues for the speaker/hearer of German, who will have to consider the articles (die Feder but der Kater) before deciding upon their gender. In Spanish, on the other hand, the noun inflections can be absolutely relied on. The suffixes -o and -a mark male and female gender.

According to Langacker’s theory, they are “noun-forming suffixes, which implies that they are themselves schematic nouns” (Langacker 2002:305). The degree of schematicity they are assigned by the speaker varies according to context. In those cases where they are attached to animate nouns they correlate with semantic gender and mean “male creature” and “female creature” ([MALE CREATURE/-o] and [FEMALE CREATURE/-a] in CG notation) but when attached to inanimate nouns such as mesa (table), they turn into highly schematic nouns whose meaning could roughly be described as “thing” (cf. Langacker 2002:306; [THING/-o] and [THING/-a]).

Thus a Spanish noun such as [FEMALE CAT/gata] is a composite structure, with the components [CAT/gat] and [FEMALE CREATURE/-a][vii]. As “[i]t is usual for the composite structure to inherit its profile from one of the components” (Langacker 2005:169), it is probably [FEMALE CREATURE/-a] which provides said structure in our case, but CG theory is very sketchy in this area. The component providing the structure also usually provides a schematic substructure, matching the profile of the other component, which is called ‘e-site’[viii] (cf. Langacker 2005:169f.). One might suggest that [FEMALE CREATURE/-a] contains a schematic substructure requiring a noun.

However, difficulties may arise, considering that gender may not be computed at all but simply learned[ix], so that the specific units are learned and memorized in an inventory together with “the constructional schemas they instantiate” (Langacker 1991:183).[x] It becomes doubtful whether it is still viable to claim that the suffixes -o and -a are meaningful. Langacker’s changes in schematicity between gata and mesa sound suspiciously like an excuse to preserve his theory’s claims intact, especially as he is keeps the way the change is worked in the mind of the speaker/hearer of Spanish unclear.

Furthermore, and more importantly, he utterly ignores the possibility that even in inanimate nouns such as cerro (hill) the suffix -o may still signify ‘male’, cognitively speaking, as witnessed by a multitude of idioms and poetic uses (cf. Jakobson 1966 on Russian grammatical gender of inanimate nouns). In order for CG to cover this possibility, too, the schematic nouns have to be regarded as polysemous to the extent that -o for example will mean [MALE CREATURE/-o] and [THING/-o] both, even when referring to a table[xi]. The notation we will hereafter use is thus [(MALE) THING/-o][xii]. Whether the ‘masculine’ aspect is processed or not depends on context (cf. Chapter 1.3.).

1.2. The extension (German)

Whereas it might be sufficient for a discussion of Spanish to restrict the analysis to the morphology of nouns, this is not the case with German where close attention needs to be paid to the articles. Although Spanish articles agree in gender with the noun they precede, just as German articles do, in Langacker’s book (2002) articles are not discussed with regard to gender but with regard to grounding. This means “that the profiled entity bears some relationship to the ground in regard to such fundamental issues as reality, [...], and speaker/hearer knowledge” (Langacker 2002:321). The la in the nominal expression la gata does not add anything semantically with regard to gender.

In a discussion of German, however, the article is more important as a marker of gender[xiii]. The ending -er, for example, can not be trusted to refer to one gender only, as it can be found both in [FEATHER/Feder] and in [TOMCAT/Kater]. Adding the article, however, solves this problem, making clear that the construction should not be [TOMCAT/Kater] but [TOMCAT/der Kater].[xiv] In this case the components would be [TOMCAT/Kater][xv] and [(MALE) THING/der][xvi]. The parentheses around ‘Male’ in the second component indicate the schematic substructure.

As in Spanish, the schematic noun [(MALE) THING/der] is a noun of varying schematicity, heavily dependent both on memorized meaning and context, as argued above. Also, the articles der (male), die (female) but not das (neuter) can apply to noun inflections such as the genitive in the case of der (cf. (1)) or plural in the case of die (cf.(2))


(1) Der Hund der Nachbarin ist krank.

(2) Die Kekse meiner Oma sind sehr lecker.

Note that the noun will be morphologically changed according to the particular requirements of the inflection in question, so that the composite structures [CATS/die Katzen] and [(FEMALE) CAT/die Katze] can be clearly distinguished. In the instance of (2), die might be simply described as [THINGS/die]. Context, obviously, must play a major role in the construal of meaning in the face of such complex networks of constructions. An excellent context theory within the framework of CG is the theory of mental spaces.

1.3. Mental Spaces

1.3.1. Literary Space

The “meaning potential” (Fauconnier 2004:661) of grammatical gender in German is manifold. In this paper, however, we will focus on the two possibilities outlined in the framework: taking grammatical gender as a “cognitively entrenched [...], necessarily schematic” (Langacker 1997:236) part of language, as far as inanimate nouns are concerned, or allowing for the cognitive processing of semantic gender, however this might work with words as die Feder.

The trouble with words such as die Feder is that common sense does not allow for the feminity of a feather, because, in the real world, a feather does not have biological gender. Yet, in Fauconnier’s theory linguistic expressions refer not to things “in the ‘real world’, but [to] things in a ‘mental space’, that is [to] a situation as conceived of by a language user.” (Taylor 2002:72). clearly, there are some situations where a language user may conceive of a feather as feminine, mostly in literary contexts.

The entity[xviii] designated by the noun (feather, for instance) is the instantiation of a ‘role’ in a certain ‘mental space’ (for the theory see Fauconnier 1984:63f.). This particular mental space will be called the literary space L, which refers not to any literary text[xix] but to a certain text. The role r1 might be ‘a friend’, ‘a servant’, or somebody else. The value r1(L)[xx], referring to the particular entity which instantiates that role in the particular space L. This space is created locally whenever the text is read and only during that time, r1(L) could, in our example, be Feather[xxi].

Yet there will be at least one other space that could be accessed by the same linguistic expression, which would be the default interpretation of the inanimate noun, it will be called the Thing Space (T). r1, obviously, would not work in this space, ‘feather’ would have to assume another role (r2) and a different value r2(T) of course. This means that depending on which space or role the linguistic expression is evoked in, its meaning is affected in direction of one of these spaces.

1.3.2. Blending Theory

The disparity of the two outlined spaces calls for two different roles and of course produces different values. However, as both sorts of spaces are needed, a slightly different approach has to be taken, such as Blending Theory (cf. Fauconnier 2004:667f.).

(3) “In [...] blending theory, [...] four mental spaces [are evoked]: two input spaces [...]; a generic space, which abstracts the commonalities from the two spaces [...] and thereby defines the cross-space mapping between the elements in the two input spaces; and a blended space, which creates a novel expressive effect” (Croft and Cruse 2004, 39; emphasis theirs)

Thus, the two disparate spaces blend in the understanding of the literary text, as both the featheriness of the feather and the gender of the person Feather are bound to be included in the understanding of the text.

2. Translation

Translation as a possible field of application for Cognitive Linguistics has been neglected for many years and even today there has little been published. Even though, as Professor Tabakowska, the foremost expert on this topic, rightly points out, CL has much to add to difficult notions such as translation equivalence[xxii] (cf. Tabakowska 1993b:73 et passim) CL concepts such as ‘imagery’ or ‘mental spaces’ could add a third level of interpretation, between a more abstract ‘meaning’ and the actual words.

The choices a translator has to make can be clarified to him by explaining the imagery that is created by certain grammatical constructions (cf. Tabakowska 1993a:796), for instance the importance of the definite article in a language such as English for a Russian translator, whose native language does not normally use the definite article except for special emphasis[xxiii].

This example is useful in demonstrating why mental spaces cannot simply be transferred, although they seem to be fairly unconnected to grammatical constructions. However, as we stressed above, it is grammatical contsructions that prompt mental spaces. To an English writer there is a difference between “seeing the owl” and “seeing Owl”. In the latter case, ‘owl’ might be a proper name. Whereas “the owl” in the former phrase would, in Fauconnier’s notation, needs only to be described as r2(T), without obvious ambiguities, the “Owl” in the second phrase can be construed both as r1(L)and as r2(T). In Russian, the word would both times be simply CABá, without an article, consequently, the ambiguity would be far more apparent.

Whether this difference in grammar mirrors a difference in construal and, ultimately, in thinking, is contested. The theory of Linguistic Relativism puts forward that very claim and, on top of that, denies the possibility of accurate translation.

3. On Linguistic Relativity

3.1. Gender

3.1.1. Grammatical and natural gender

English and German differ fundamentally in respect to gender, mainly because English does not have grammatical gender: “Gender in English is primarily a semantic category” (Hellinger 2001:107), which means, that for the most part, only those nouns are gendered that refer to a human whose gender is known. Yet this rule is not dependable, as there are not only exceptions such as social gender[xxiv] (cf. Hellinger 1990:61) or psychological gender[xxv] but it also seems that speakers of English “have a ‘folk theory’ of gender[xxvi] that enables them to assign gender to objects that lack natural gender” (Sera, Berge and Pintado 1994:287). Furthermore, it has been frequently remarked that in American discourse, “reference to women [...] was much less frequent than reference to males” (Romaine 2001:161).[xxvii]

In German the issue is more straightforward, as every single noun is assigned a grammatical gender. Once assigned, it is usually fixed (cf. Bußmann and Hellinger 2003:145). Gender assignment is often taken to be arbitrary, yet “[f]or approximately 90% of German monosyllabic nouns, gender class membership can be predicted” (143). Moreover, some of the criteria for classification in English ‘folk theory’[xxviii] seem to be at work in grammatical gender assignment, too (cf. Sera, Berge and Pintado 1994:288).

Conceptually, we can describe these different criteria for gender assignment easily with the mental spaces vocabulary, recalling that linguistic expressions do not necessarily refer to ‘the real world’ but to mental spaces. One of the reasons why gender assignment in one’s own language is rarely questioned[xxix], might be that grammatical gender refers to a so-called base space, which is a default space (cf. Taylor 2002:72f.), representing “the mutually known world of the interlocutors” (Croft and Cruse 2004:33). Yet what people ‘know’ about gender, they learn early on, by learning their language (cf. Howard 2001:191).

3.1.2. Learning Gender

The question how gender is processed, especially by children, and the importance of the category of gender for the child’s understanding of the world is vital for the theory of Linguistic Relativity and hence has to be raised in this paper.

Until children are 2 years old, they assign gender randomly. Yet older children show a strong bias towards gender stereotyping (cf. Liben and Signorella 1993). There is strong evidence that children are “invited” (cf. Waxman and Markow 1995) by words to form categories, or at least they are alerted by words or labels to the existence of such categories (cf. Waxman and Gelman 1986; see also Spelke and Tsivkin 2001) early on in their development.

Grammatical gender is learned fairly early, compared to semantic gender (cf. Taylor 1989:247), and it takes children quite some time to form “stable shared conceptual systems” (Lucy and Gaskins 2001:280). During this time children possibly learn by learning language “that objects belong in categories” (Gopnik 2001:60) and as they “assume that giving the same name to two objects means that they share a common nature” (61), they might be predisposed to gender stereotyping by having been exposed to the linguistic category of gender.

In this context, it becomes highly significant that personalization in German children’s literature corresponds not only to grammatical gender but also to gender ‘folk theories’ (cf. Corbett 1991:94f.). If children learn their gender stereotypes through the books they are read as toddlers, it makes sense to compare two versions of one of the most famous children’s book, Winnie-the-Pooh.

3.2. Translation

3.2.1. Vertical and Horizontal Translations

The two versions about to be compared are not two drafts of the text but the ‘original’ text and the translated German version of it. This comparison begs the question of translatability. If a text can be translated easily, without losing anything, then any difference between the two versions must be due to bad translating skills. Even though the Sapir-Whorf-Hypothesis, which entailed the impossibility of translation, has not been without influence in philology, within translation studies, there has never been particularly much interest in the hypothesis.

Instead, much effort has been invested into showing what a faithful translation might look like, by concetrating mostly on the idea of ‘vertical’ translation, which views translation “as containing two main processes: full comprehension of the [...] text [...], followed by production of the constructed meaning” (De Groot 1997:30). The exact phrasing, the grammatical and lexical aspects of the text get lost. Yet, as is often claimed, this technique is necessary for literary translation[xxx].

Yet, in view of the previous discussion of gender, the importance of a horizontal view of translation, which “construes translation as ‘transcoding’” (30) cannot be underestimated, as only in this way can grammatical structures be taken into account.

3.2.2. Translatability

The answers to the question of translatability revolve around the notion of commensurability, which refers mostly to the commensurability of the two conceptual systems (of the two languages at either end of the translation process). Understanding a text in a different language requires only few “correspondences in well structured experiences and a common conceptualizing capacity” (Lakoff 1987:312). Yet even that can theoretically be problematic, as Whorf argues:

(4) “A scientist from another culture that used time and velocity [instead of time and space] would have great difficulty in getting us to understand these concepts.” (Whorf 1956:218)

This is merely about understanding, not yet about translation. For a ‘good’ translation one requires “close correspondences across conceptual systems” (Lakoff 1987:312). Thus, a translation from English into Hopi or Yukatán might be difficult, with much being lost. This, however, need not concern us here, as our two languages are closely related, as they share many aspects, down to covert categories. Nevertheless, whether gender is translatable from English to German remains to be seen.

4. Translating Winnie-the-Pooh

4.1. Gendering Owl (English)

This section of the paper is concerned with The Complete Winnie-the-Pooh and the depiction of gender within it, specifically of Owl’s gender. The characters of the book, Piglet, Winnie-the-Pooh, Eeyore, Owl and others seem at first glance to be removed from the sphere of human gender concerns. Yet, the book does not only contain a human character, Christopher Robin, there is also an explicit concern with gender, which is made clear in the opening pages:


(5) When I first heard [Winnie-the-Pooh's] name, I said, just as you are going to say, “But I thought he was a boy?”
“So did I,” said Christopher Robin.
“Then you can’t call him Winnie?”
“I don’t.”
“Butyou said-”
“He’s Winnie-ther-Pooh. Don’t you know what ‘ther’ means?”
(Milne 2002:15)

For this paper, taking the eponymous bear as a focus would not have made much sense, we will later see why. Nevertheless, this passage shows that gender does constitute a concern in this text, or at least the text is not oblivious to it.

Owl is a male character. His gender is marked only by the pronoun ‘he’, which is used but rarely, usually he is called or referred to by his proper name, Owl. Which also is a description of his person, as he is an owl in most respects, there is no textual evidence that he is wearing something similar to human clothes. His human characteristics consist of him living in a tree-apartment, with a proper door and signs and -for a short time, anyway- a doorbell, and, of course of his speaking English.

Linguistically, Owl is an intersting case, as his name also refers to his biological nature. Owls, however, do not normally have a gender in English, so that Owl’s gender is an invention of the author[xxxi]. He could have had a female Owl as well, it seems. Yet, recalling English ‘folk theories’, it is probable that there is an intuition for native speakers of English to categorize owls als male, as they tend to characterize bears and elephants as male (cf. Corbett 1991:95). Linguistically, Owl is only marked by the missing article[xxxii]. A CG description as in Chapter 1.1. is out of the question.

Luckily, we also have the mental spaces theory. As a grammatical prompt for a mental space, the missing article and the gendered pronouns are sufficient to assign Owl to a role r, which could be described as ‘avuncular, wise friend’ in the literary space L[xxxiii]. The value r(L) would then be Owl.

4.2. Translating Owl (German)

In the translated version, the translation of the character Owl seems to have been pretty straightforward. He is called Eule, a proper name and a biological refernce at the same time. However, checking the gender of Eule, one notices that Eule is female.[xxxiv]

At first glance this seems to be simply a bad translation: Harry Rowohlt jumbled the genders. Yet at second glance this judgement is less convincing. In German, the owl, die Eule, has a female grammatical gender. Whereas in English Milne was, at least from a grammatical point of view, free to assign any gender he wanted to without sounding especially odd. Assigning Eule a male gender would have a strange ring to it.

Comparing the two texts linguistically, it will be noticed that as far as mental spaces are concerned, in German a second space is activated, which we will call the Animal Space (A). A female Eule is so much in accordance with its biological companion die Eule, that the role r2, which might be described as ‘flying animal which eats field mice” in space A, where die Eule is the value of r2(A). This unwieldy description can be evaded by creating a blending space. The input spaces would be spaces L and A and the generic space would contain feathers, a home in the trees etc.

If we take a look at the way Eule is linguistically described in the framework put forward in chapter 1.2., we can see why she would prompt this kind of mental space. In view of the fact that [OWL/Eule] is not a self-sufficient noun, that it is a component which needs the other component [(FEMALE) THING/die] in order to form a properly functioning composite structure, it becomes obvious why the Animal Space is triggered by the proper name Eule. However, would Eule have had a male gender, the distancing effect created by that deviation might be sufficient to tone down the nouniness of the word/name Eule.

Would this, then, have been a better translation? This is a hard question to answer, as there is a third aspect to consider, the intuition of the native speaker and ‘folk theories’ concerning gender. As these theories mostly closely follow grammatical gender, the German reader might be put off in a way that the English reader is not. It was said earlier in this paper, that the folk theories resemble each other closely, yet they are not identical.

The question of the right choice has to be, in the end, a pragmatic decision. There are risks both ways.

5. Conclusion

Whorf famously demanded of Western Knowledge “a re-examination of the linguistic background of its thinking” (Whorf 1956:247). What we have attempted to in the present paper, on a very small scale, is to examine the Cognitive Linguistic background of gender and translation, with the Whorfian question of translatability as theoretical backdrop.

What about thought and language? This question has always been difficult to answer and an answer will not be attempted here. However, the influence of language on gender as a category and on translation has been suggested. Especially the latter suggestion recalls a recent reformulation of the Whorfian Hypothesis, i.e. Slobin’s formula of “thinking for speaking” (Slobin 1990:75). We do not know how the mind would work in absence of language, but when ‘thinking for translating’, language almost certainly exerts its influence. A native speaker of Spanish, with possibly different ‘folk theories’ and different grammatical genders, were he called upon to translate the text from English into German, would he have made the same choice as Harry Rowohlt?

What about the children? Does reading Pu der Bär influence children’s stereotyping differently than Winnie-the-Pooh does? There are feminist translators, who certainly believe that, changing as they do sexist gender typings whenever translating a text (cf. Von Flotow 1997:24ff.). The present paper did not look for evidence and so it did not provide any, yet it did show which alleys one might pursue, were one to look for such evidence.

This paper has one big omission, cognitive semantics, because this would have been by far to big a subject for the scope of this paper. The recognition of the importance of inflections for cognitive processes may lead to a reformulation of cognitive semantics which includes inflections[xxxv], and not just gender inflections at that.

Then, maybe, will a truly multidisciplinary gender study be possible, as Pütz, one of the founders of Critical Cognitive Linguistics envisions it (cf. Pütz 2005:155f.).

Lastly, we return to the only question not answered: the question of translatability. True equivalence is hard to attain, if not impossible, that seems clear after trying to translate the simple word ‘Owl’ led to an ambivalent situation. Translation is only ‘possible’ if one takes a highly pragmatic view of translation.

6. Bibliography

Baron, Dennis. 1986. Grammar and gender. New Haven: Yale University Press

Bickel, Balthasar. 2000. Grammar and social practice: On the role of ‘culture’ in linguistic relativity. Evidence for linguistic relativity, ed. Susanne Niemeier and René Dirven, 161-192. Amsterdam: Benjamins

Bußmann, Hadumot and Marlis Hellinger. 2003. Engendering female visibility in German. Gender across languages: The linguistic representation of women and men, Vol.3, ed. Marlis Hellinger and Hadumot Bußmann, 141-175. Amsterdam: Benjamins.

Corbett, Greville. 1991. Gender. Cambridge: Cambridge University Press.

Croft, William and D. Alan Cruse. 2004. Cognitive linguistics. Cambridge: Cambridge University Press.

De Groot, Annette M. B. 1997. The cognitive study of translation and interpretation: Three approaches. Cognitive processes in translation and interpreting, ed. Joseph H. Danks et al, 25-57. Thousand Oaks: Sage.

Fauconnier, Gilles. 1984. Espaces mentaux: aspects de construction du sens dans les langues naturelles. Paris: Minuit,1997.

,– Mappings in thought and language. Cambridge: Cambridge University Press. 2004. Pragmatics and Cognitive Linguistics. The handbook of pragmatics, ed. Laurence R. Horn and Gregory Ward. Malden: Blackwell.

Fillmore, Charles J. 1977. Scenes-and frames-semantics. Linguistic structures processing, ed. Antonio Zampolli, 55-81. Amsterdam: North-Holland.

Gopnik, Alison. 2001. Theories, language and culture: Whorf without wincing. Language Acquisition and conceptual development, ed. Melissa Bowerman and Stephen C. Levinson, 45-70. Cambridge: Cambridge University Press.

Hellinger, Marlis. 1990. Kontrastive feministische Linguistik: Mechanismen sprachlicher Diskriminierung im Englischen und Deutschen. Ismaning: Hueber. 2001. English – gender in a global language. Gender across languages: The linguistic representation of women and men, Vol.1, ed. Marlis Hellinger and Hadumot Bußmann, 105-114. Amsterdam: Benjamins.

Howard, Henry. 2001. Age/Gender morphemes inherit the biases of their underlying dimensions. Language and Ideology: Theoretical Cognitive Approaches, ed. René Dirven et al, 165-196. Amsterdam: Benjamins.

Hyde, G. M. 1993.The Whorf-Sapir Hypothesis and the translation muddle. Translation and Literature 2: 3-16.

Jackendoff, Ray. 2002. Foundations of language: Brain, meaning, grammar, evolution. Oxford: Oxford University Press.

Jakobson, Roman. 1966. On linguistic aspects of translation. On Translation, ed. Reuben A. Brower, 232-240. New York: Oxford University Press.

Köpcke, Klaus-Michael and David A. Zubin. 1984. Sechs Prinzipien für die Genuszuweisung im Deutschen: Ein Beitrag zur natürlichen Klassifikation. Linguistische Berichte 93: 26-50.

Lakoff, George. 1987. Women, fire and dangerous things: What categories reveal about the mind. Chicago: University of Chicago Press.

Langacker, Ronald W. 1987. Foundations of cognitive grammar, vol 1: Theoretical prerequisites. Stanford: Stanford University Press.

.– 1991. Foundations of cognitive grammar, vol2: Descriptive application. Stanford: Stanford University Press.

.– 1997. The contextual basis of cognitive semantics. Language and conceptualization, ed. Jan Nuyts and Eric Pederson, 229-252. Cambridge: Cambridge University Press.

.– 2002. Concept, Image, and Symbol: The Cognitive Basis of Grammar. 2nd Ed. Berlin: de Gruyter.

.– 2005. Integration, grammaticization and constructional meaning. Grammatical constructions: back to the roots, ed. Mirjam Fried and Hans C. Boas, 157-189. Amsterdam: Benjamins.

Leino, Jaakko. 2005. Frames, profiles and constructions: Two collaborating CG’s meet the Finnish permissive construction. Construction grammars: Cognitive grounding and theoretical extensions, ed. Jan-Ola Östman and Mirjam Fried, 89-120. Amsterdam/Philadelphia: Benjamins.

Liben, Lynn. S. and Margaret.L. Signorella. 1993. Gender- schematic processing in children: the role of initial interpretations of stimuli. Developmental Psychology 29: 141-149.

Lucy, John A. 1992. Language diversity and thought: A reformulation of the linguistic relativity hypothesis. Cambridge: Cambridge University Press.

.–1997. The scope of linguistic relativity: an analysis and review of empirical research. Rethinking linguistic relativity, ed. John J. Gumperz and Stephen C. Levinson, 37-70. Cambridge: Cambridge University Press.

Lucy, John A. and Suzanne Gaskins. 2001. Grammatical categories and the development of classification preferences: A comparative approach. Language acquisition and conceptual development, ed. Melissa Bowerman and Stephen C. Levinson, 257-284. Cambridge: Cambridge University Press.

Milne, A.A. 1998. Pu der Bär. Transl. Harry Rowohlt. Hamburg: Dressler.

.–1999. Pu baut ein Haus. Transl. Harry Rowohlt. Hamburg: Dressler.

.– 2002. The complete Winnie-the-Pooh: containing Winnie-the-Pooh and The House at Pooh Corner. London: Dean.

Nesset, Tore. 2001. How pervasive are sexist ideologies in grammar? Language and Ideology: Theoretical Cognitive Approaches, ed. René Dirven et al, 197-227. Amsterdam: Benjamins.

Nissen, Uwe K. 2002. Gender in Spanish: Tradition and Innovation. Gender across languages: The linguistic representation of women and men, Vol.2, ed. Marlis Hellinger and Hadumot Bußmann, 251-281. Amsterdam: Benjamins.

Pederson, Eric. 1995. Language as context, language as means: Spacial cognition and habitual language use. Cognitive Linguistics 6: 33-62.

Pütz, Martin. 2005. Diskurs, Ideologie und Metaphern: Aspekte einer kritischen kognitiven Linguistik. Texte: Spielräume interpretativer Näherung, ed. Stephan Merten and Inge Pohl, 147-171. Landau: Knecht.

Romaine, Suzanne. 2001. A Corpus-based view of gender in British

and American English. Gender across languages: The linguistic representation of women and men, ed. Marlis Hellinger and Hadumot Bußmann,153-177. Amsterdam:Benjamins.

Scarpa, Federica. 2002. Closer and closer apart? Specialized translation in a cognitive perspective. Translation Studies: perspectives on an emerging discipline, ed. Alessandra Riccardi, 133-149. Cambridge: Cambridge University Press.

Schwink, Frederick W. 2004. The Third Gender: Studies in the origin and history of germanic grammatical gender. Heidelberg: Winter.

Sera, Maria D. and Christian A. H. Berge and Javier Del Castillo Pintado. 1994. Grammatical and conceptual forces in the attribution of gender by English and Spanish speakers. Cognitive Development 9: 261-292.

Slobin, Dan I. 1996. From “Thought and Language” to “Thinking for Speaking”. Rethinking linguistic relativity, ed. John J. Gumperz and Stephen C. Levinson, 70-96. Cambridge: Cambridge University Press.

Spelke, Elizabeth S. and Sanna Tsivkin. 2001. Initial knowledge and conceptual change: Space and number. Language Acquisition and conceptual development, ed. Melissa Bowerman and Stephen C. Levinson, 70-100. Cambridge: Cambridge University Press.

Tabakowska, Elźbieta. 1993a. Articles in translation: an exercise in cognitive linguistics. Conceptualizations and mental processing in language, ed. Richard A. Geiger and Brigida Rudzka-Ostyn, 785-801. Berlin: De Gruyter.

.– 1993b. Cognitive Linguistics and poetics of translation. Tübingen: Narr.

Taylor, John R. 1989. Linguistic categorization: Prototypes in linguistic theory. Oxford: Claredon.

.– 2002. Cognitive Grammar. Oxford: Oxford University Press.

Van Berkum, Johannes J. A. 1996. The psycholinguistics of grammatical gender: studies in language comprehension and production. Nijmegen: Nijmegen University Press.

Von Flotow, Louise. 1997. Translation and Gender: translating in the era of Feminism. Manchester: St. Jerome.

Waxman, Sandra and Rochel Gellman. 1986. Pre-schoolers’ use of superordinate relations in classification and language. Cognitive Development 1: 139-156.

Waxman, Sandra R. and Dana B. Markow. 1995. Words as invitations to form categories: Evidence from 12-13 month-old infants. Cognitive Psychology 29: 257-302.

Whorf, Benjamin Lee. 1956. Language, thought and reality: selected writings of Benjamin Lee Whorf, ed John B. Carroll. Cambridge: The M.I.T. Press.


[i] For a brief but clear discussion of the relationship between CxG and Scenes-and-Frames-Semantics see Leino 2005.

[ii] Cf. Taylor 2002 for an easy but cursory introduction and Langacker 2002 for a more difficult yet more rewarding treatment.

[iii] It has to be stated that Nesset’s paper is very speculative, as he overcrowds his paper with feminist assumptions (drawn mostly from Beauvoir), which tend to muddle the clarity of his analysis. For a clearer (though less spectacular) paper on the same topic with comparable results see Howard 2001.

[iv] An author, who, interestingly, has discussed, as early as 1940, the intricacies of male and female pronouns (cf. Baron 1986:213).

[v] To translate is to betray.

[vi] For a brief survey of Spanish gender see Nissen 2002.

[vii] The components for [TABLE/mesa] would be, of course, [TABLE/mes] and

[THING/-a].

[viii] Short for ‘elaboration site’.

[ix] Although rules and lists do coexist in CG (cf. Langacker 1987:42).

[x] This claim that gender is not computed but memorized and stored in an inventory has since been significantly substantiated by evidence from anomic patients (cf. Van Berkum 1996:186n and Jackendoff 2002:215n).

[xi] Langacker does regard -o and -a as polysemous but only until they are attached to a noun. Then they seem to take on a single meaning (cf. Langacker 1991:185).

[xii] Notice that we did not lose the ‘creature’ part of the description; it is contained in the ‘(MALE)’ part. A difference between creatures or things needs no longer to be entered in the description.

[xiii] However, this does not mean that the definitive article has lost its grounding function, but, compared to Spanish, the function as a gender marker is strongly foregrounded.

[xiv] Cf. also: “By asking what gender a noun like German Tisch ‘table’ has, we may be missing the underlying form in the lexicon. The appropriate question may very well be ‘What is the gender of der Tisch?’” (Schwink 2004:5)

[xv] Of course, this notation of the component parts might be seen as being contradictory, yet it is not. It is just that [TOMCAT/Kater] is the only way to describe this component, you can’t put the ‘TOM’ in parentheses.

[xvi] In analogy to our notation concerning suffixes in the previous section.

[xvii] This example is difficult because cat as well as die Katze refers both to an individual cat of unspecified gender as well as to a female cat.

[xviii] The ontological status of this entity is unclear. It is not ‘in the world’ but it is not’in a mental space’ either.

[xix] “Literary text” does not mean here the material text ‘in the real world’ but the text as it is read and understood by one individual .

[xx] In Fauconnier’s theory this works like a mathematical function.

[xxi] The capital letter indicates that the word ‘feather’ might in this case be used as a proper name.

[xxii] A different angle than in Tabakowskas work is pursued in Scarpa 2002, where the focus is on specialized literature.

[xxiii] That means it is more like a demonstrative than like a definite article.

[xxiv] Gender determination through “Eigenschaften des prototypischen Referenten” (Hellinger 1990, 65).

[xxv] “affective attitudes of the speaker, which accounts for variation as in babyshe vs. babyit” (Hellinger 2001, 107f.).

[xxvi] This is actually reminiscent of an interesting point of Bickel, who claimed that whoever wanted to explain Linguistic Relativity had to take sociocultural practice into account, too (cf. Bickel 200).

[xxvii] Compare Whorf’s fascinating idea that covert grammatical categories such as the English gender system “are more likely to be ‘rational’, [...] in accordance with [...] nonlinguistic fact” (Lucy 1992:28).

[xxviii] Most importantly the artificial-male/natural-female division (cf. Sera, Berge and Pintado 1994:287).

[xxix] Cf. Whorf 1956:207ff.

[xxx] Many of these strange arguments are provided in Hyde 1993, an attack on the Sapir-Whorf-Hypothesis.

[xxxi] And one of the human features of Owl.

[xxxii] The capital ‘O’ does not count, as the book has a very sloppy orthography as far as capital letters are concerned.

[xxxiii] Again, it has to be stressed, that these spaces are not ‘there’, they imply someone reading this text at this moment.

[xxxiv] Notice that Pu der Bär is a male character, just as in the original.

[xxxv] One of the most comprehensive works on cognitive semantics, Leonard Talmy’s Towards a Cognitive Semantics mostly ignores inflections.