Excerpt
Contents
1 Introduction
2 The relation between language and thought
2.1 The effect of language on thought
2.2 Linguistic relativity - a definition and discussion
3 The grammatical gender and its impact on thought
3.1 Grammatical gender - a definition
3.2 Empirical research on grammatical gender and its effect on the categorization of inanimate objects
4 Conclusion
1 Introduction
Throughout history, the relation between language and thought has been of interest for scholars of different fields such as philosophy, psychology and linguistics. Whereas Wittgenstein was convinced that language is thought, as “[n]ow it is becoming clear why [he] thought that thinking and language were the same. For thinking is a kind of language,” (1969: 82) others like Pinker (1994) and Fodor (1975) believed there is no relation whatsoever and that thought is independent of language. However, with regard to studies confirming the notion of language indeed having an effect on cognition, the broad consensus is that neither of these extreme approaches is correct and that the truth is situated somewhere in between.
In this context, it is the Sapir-Whorf (or Whorfian) hypothesis-better known as the linguistic relativity hypothesis-that has gained great attention amongst psycholinguists, for at its heart it holds the very idea that the languages we speak at least influence the way we think. Evidence supporting this theory stems from various domains describing differences between languages with respect to perceptions and judgments of concepts, such as quantity, space and time.
In recent decades, however, the grammatical gender of nouns in particular, as one aspect of grammar, has been the focus of investigations and debates on linguistic relativity. Several cross-linguistic studies indicate that speakers of languages which have a gender system tend to assign stereotypical attributes and properties to inanimate objects congruent with their grammatical gender. Though there is empirical evidence substantiating this hypothesis, there is a considerable amount of research findings which show little to no support, which therefore leads to the question whether grammatical gender is a useful tool for the investigation of linguistic relativity after all.
By this token, the present paper will provide a collection of works which outline this discussion, showing that certain aspects need to be taken into consideration when assessing the grammatical gender as evidence for language influencing thought. The choice of selection is based on the fact that their research findings build up on each other’s content and allow transitions in a logical manner. Definitions and elaborations of linguistic relativity and grammatical gender in the beginning and a recently published review of empirical studies as an outlook for future research frame this paper.
2 The relation between language and thought
2.1 The effect of language on thought
Focussing on the relation between language and thought, linguist and psychologist John A. Lucy generally differentiates between three kinds of effects: The first one is called the semiotic level and refers to the question of whether or not the human species has a symbolic language, or more precisely, if having a code which is equipped with a constituent part of symbolic nature is responsible for the transformation of thoughts (1997: 292). Interrelatedly, it raises the question whether it makes a difference on how we think versus how other species think who do not have symbolic languages (cf. ibid.).
The second type of language effect is structural and relates to the differences among the languages with respect to their grammar, vocabulary etc. It constitutes a domain-centered approach and deals in general with the question of whether and how different linguistic systems with distinct morphosyntactic formations of meaning affect the way of thinking about the perceived and experienced reality. In that matter, it is a case of linguistic relativity (cf. ibid.).
If, however, peoples’ encoding of the reality is influenced by the ways of using language, Lucy argues the effect of language on thought is functional. In that regard, it is a behavior-centered approach, and revolving around the differences in uses of language, it raises the question if a specific type of register in an academic or professional context, for instance, affects interactional, interpretive or analytical aspects in our processes of thought (cf. ibid.).
The second type of effect constitutes the essence of the discussion on grammatical gender and lays the foundation for the investigations in the following chapters.
2.2 Linguistic relativity- a definition and discussion
Having its origins in ancient times already, the principle of linguistic relativity is attributed to linguist Benjamin Lee Whorf and his Ph.D. advisor Edward Sapir, a renowned anthropologistlinguist who laid the groundwork in that matter (cf. Kurinski & Sera: 203). Whorf (1956)’s field work on the Uto-Aztecan language Hopi spoken in Southern Arizona and finding out that their language lacks the concept of tense which indicates that inferences about time have to be made through other sentential contexts, marks the beginning of his formulation of the theory. He was convinced that language influences thought - an idea which eventually has come to be known as linguistic relativity. Whorf argued that, “[i]t was found that the background linguistic system (in other words, the grammar) of each language is not merely a reproducing instrument for voicing ideas but rather itself the shaper of ideas, the program and the guide for the individual’s mental activitiy” (1956: 212). This theory established the notion that thought is constrained by language and was regarded as linguistic determinism.
In a wider sense, however, this theory points out that linguistically established habits have a crucial impact on forming non-linguistic behavioral patterns, which is also supported by Sapir who claims that “[...] language habits of our community predispose certain choices of interpretation” (Sapir 1949: 162). Though Whorf himself never made the claim that thought was the one and only determining factor as he noted that “[he] [doesn’t] wish to imply that language is the sole factor in types of behavior [...], but that this is simply a coordinate factor along with others [...]” (Lee 1996: 153), his theory was linked to the notion of determinism for a long time and thus, subject to much controversy in this regard. Nevertheless, one could argue that what Whorf mainly suggests is that the use of language produces and/or reinforces habits when deciding how to behave in certain contexts, for “[w]e dissect nature a long lines laid down by our native language [...] the world is presented in a kaleidoscopic flux of impressions which has to be organized by our minds - and this means largely by the linguistic system in our minds” (1956: 213).
Eventually, linguistic determinism as well as its counterpart, the “universalist” position which states that thought is independent of language, have been disproven (cf. Samuel et al: 1767). However, Whorf’s idea of linguistic relativity finds recognition in Slobin (1996)’s concept of thinking for speaking which constitutes linguistic relativity as the mental activity requiring a person to pay attention to the distinctions obligatory expressed in the respective language and to verbalize different aspects of reality when constructing linguistic outcome in the process of using any language. In other words, thinking for (or about) speaking “might allow language to bias our attention toward the more linguistically describable aspects of what we perceive [...], but this does not mean that language alters the underlying conceptual structures” (Samuel et al. 2019: 1767).
For that matter, Whorf’s linguistic relativity hypothesis, which was still relevant and of interest in the fields of research, was differentiated into two versions: the strong version, which referring to linguistic determinism, was discounted and the weak version that in essence states that there is some kind of influence, albeit language is not the root of thought.
Cross-linguistic comparisons indeed show evidence for language influencing thought in various domains, such as space/movement (e.g. Levinson et al. 2002), time (Boroditsky 2011) and matter (Lucy 1992). With respect to space and movement for instance, there are languages like Tzeltal (a Mayan language spoken in the Mexican state of Chiapas) which use an absolute directional or coordinate system “according to which the positioning of referents are described with reference to some physical phenomena in the environment, typically the "movement" of the sun” (Everett: 67). In contrast, English makes use of a relative directional system, in which the reference points for orientation vary and thus, must be mediated otherwise within the communicational context (e.g. one’s left side might me the right side of another’s).
Overall, current investigations do no longer put the focus on the question, whether language affects thought, as it is greatly agreed upon that it does, but rather in which ways language does and does not relate to it and what the consequences of linguistic relativity are with respect to cognition. One hypothesis for instance states that the linguistic classifications imposed by language only have an effect on other linguistic processes and representations (cf. Kousta et al.: 843). Another argues that “language also affects nonlinguistic cognition, above and beyond the processes involved in speaking and comprehending a particular language. In other words, the main debate in the field focuses on the extent and pervasiveness of the effect of language on cognition” (ibid.). These assumptions, ultimately open up a wide field for research.
3 The grammatical gender and its impact on thought
3.1 Grammatical gender - a definition
Generally speaking, grammatical gender refers to a group of classes which are assigned to all kinds of nouns. Its classification is indicated by either the form of the noun itself or by the choice of words which modify, substitute or in any other way are in reference with the noun. Unlike English, many languages have a gender system and most commonly known are the groups which consist of two, masculine and feminine (as in French or Spanish) or three genders masculine, feminine and neuter (like in German). Though different languages may vary with respect to the number of genders (from two to over twenty), in most cases the attribution of class is in part in accordance with the biological sex or animateness (Phillips & Boroditsky 2003: 928). However, there are instances in which the assignment of a gender to a noun seems to be arbitrary, for in some languages the genders for the same object are different. Could the grammatical genders assigned to objects by a language influence people’s mental representations of objects?
3.2 Empirical research on grammatical gender and its effect on categorization of inanimate objects
As they provide the basis for investigation, a series of experiments and observations by Phillips and Boroditsky (2003) will serve as the starting point in their attempt to find out to which extent grammatical gender influences the perception, representation and categorization of inanimate objects on the cognitive level.
The first experiment was conducted to see if speakers would rate same-gender pairs which are concordant with the gender system of the language they consider their native language as more similar. Hence, Spanish-English bilinguals and German-English bilinguals (ages 17 to 69) were asked to rate the similarity of objects and animals to human males and females on a scale from 1 to 9 (9 being the number for greatest similarity). All instructions were given in English -a language in which both groups were highly proficient - and all items were presented as unlabeled pictures having opposite genders in the respective languages (for instance ‘toaster’ is masculine in German, but feminine in Spanish). The pictures referring to persons included a ballerina, a girl, a woman and a bride in the female group and a giant, a king, a boy and a man in the male one. The results showed that the subjects indeed found greater similarity between people and objects of matching gender, supporting the theory that speakers think about objects as more similar to biological males and females, when they have experience with a (native) language in which the grammatical gender is the same.
On the basis that grammatical gender appears to bias one’s representation of objects, Phillips and Boroditsky then wondered, what would happen if a person spoke two languages that gave opposite grammatical genders to an object, and if the discovery that that grammatical genders can vary would lead people to discount these first-learned grammatical gender biases. In order to find answers to these questions, they repeated the same similarity task with people who spoke both Spanish and German. Interestingly, they found a significant positive correlation between people’s relative proficiency in the respective languages Spanish and German and their biases in the similarity task, meaning the more relatively proficient a person was in Spanish for instance, the greater was the consistency of their rating with the Spanish grammatical gender. Further, in order to examine the speakers’ non-linguistic representation of objects, the experiment was repeated a second time with the implementation of a verbal interference task, making it difficult for them from to name the objects subvocally. However, this too did not change the results.
Eventually, one question still remained: namely if such differences in similarity were solely rooted in differences in grammar, or if they might as well be the product of cultural differences. In order to examine this question, the researchers had to design a task which would rule out the influence of any cultural factors. Hence, they invented a fictional language which they called Gumbuzi; this fictional language followed the rule of differentiating between the two genders correspondent to biological genders, but the terms masculine and feminine were replaced by soupative and oosative. The participants were native English speakers divided into two groups (English because it lacks grammatical gender and therefore is free of biases) and they had to memorize supposedly random words which were assigned to one of the two categories; for instance, they had to learn that “pans, forks, pencils, ballerinas and girls are soupative, while pots, spoons, pens, giants, and boys are oosative” (2003: 931). After making sure that they had mastered memorizing them by testing their knowledge repeatedly, they eventually had to perform the task of rating the similarity of person-object pairs. One group had to do so while performing a verbal interference task, the other did not.
The results were congruent with the ones before; English speakers behaved just like the ones in Spanish and Germans and, they did so in both groups.
In sum, the overall results suggest that there is indeed an effect of grammatical gender on people’s ideas about the genders of objects and that the reason for this might, as Phillips and Boroditsky argue, lie in the fact that languages invite their speakers to (not necessarily consciously) carry out such comparisons, which they would not have carried out otherwise. Thus, they consider the possibility that people deliberately look for similarities between entities that are assigned to the same grammatical category, and in the process even discover meaningful similarities. This understanding of grammatical gender as a “convenient tool” leads to the assumption that grammatical gender functions like any other categorical term (such as penguin, game etc.) (cf. 2003: 932). However, reflecting on the results, one has to question if the tasks were truly non-linguistic, even though a verbal interference task was implemented; the instructions and names were still of linguistic nature. Thus, Phillips and Boroditsky’s research does not answer the question if grammatical gender can truly alter non- linguistic representations. Yet, they draw attention to another aspect which was not part of their study but that might be of great relevance with respect to linguistic relativity; in their works, they tested bilingual speakers who had experiences and equal levels of high proficiency in both of their languages. In addition, English speakers were tested with the use of a fictional language with respect to the influence of cultural differences. In this regard, one may ask the question, if it makes a difference when a second language with a different gender system is learned later on in life. If one were to suppose, in the realm of linguistic relativity, that languages affect thinking, it would be interesting to see if learning a new gender system would cause a cognitive change.
[...]