«Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 1 Sex and Size: The Influence of Grammatical Gender on ...»
Another related problem is that nouns with biological gender belong to the same category as objects with no biological gender. It seems logical that if a German-speaking subject is asked whether ‘door’ is more feminine or more masculine, having nothing else to factor into the Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 26 decision, the subject would make the selection based on grammatical gender and choose feminine.
Testing how feminine or masculine words are thought of by German speakers is difficult.
If, because of task design, subjects are able to adopt a conscious strategy for responding to the target words, then it is possible that the grammatical gender of the noun class will weigh into the subjects’ decisions about the words. Though the ability to categorize words according to categories of grammatical gender certainly indicates knowledge that German speakers can bring to bear in thinking about words given the right conditions, the question becomes whether or not this knowledge is as available and as salient as other aspects related to the semantic representation of the word.
This study attempts to account for the possibility that German-speaking subjects might create a strategy for their responses that is based purely on the grammatical gender, which is, especially in German, often based on morphology and not on semantic or cultural notions of masculinity and femininity. Another novel aspect of this study is that it examines diminutives in both English and German. As discussed above, diminutives are neither masculine nor feminine in either language, as English does not mark gender, and in German the diminutive suffix conveys neuter gender. However, in both English-speaking and German-speaking Western cultures, the diminutive has feminine associations. For this reason, we hypothesized that speakers of both languages might perceive diminutiveness as more feminine than the nouns from which they were derived.
The following experiments examine how the disparities between grammatical gender assignment and biological and semantic7 gender assignment are rectified. I will aim to give Here ‘semantic gender’ is used as Köpcke and Zubin use it to refer to groups of words with similar meanings that all have the same gender.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 27 supporting evidence that both cultural associations and grammatical gender are important to perception. Is it the gender of the word or the cultural associations of a word that influences its perception? As we have seen with the Mills study, the perception of gender varies across languages and it has been suggested that the presence of a linguistic gender system fueled that difference. Let’s now look at how cultural similarities affect German and English speakers’ perceptions of words for objects and animals, with particular attention to diminutives, a crosscultural phenomenon. It can be said that diminutives are associated with women because of their cultural attributions of smallness and innocence.
So how do we reconcile the differences between Boroditsky et al.’s findings that grammatical gender influences object perception and my friend’s observations that experiences with objects affects their meaning? A word like der Rock ‘the skirt’ is masculine in German, but clearly has an association with femininity. For a more complex investigation, I examine diminutives, which are all neuter, but may be associated with femininity because of their marking for smallness.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 28 Perceptions of German and English diminutives and non-diminutives
6.0 Objectives As discussed earlier, German uses gendered pronouns to refer to nouns. Feminine and masculine pronouns are regularly used for things that would be referred to simply as ‘it’ for English speakers. It seems plausible that over time German speakers may begin to associate ‘femininity’ with feminine objects and ‘masculinity’ with masculine objects. Since English lacks such a gender system for its nouns, English speakers may be more likely to think of all objects as ‘its’, whereas German speakers will divide them up into ‘hes,’ ‘shes’ and ‘its.’ This study was composed of two tasks conducted in German with native German speakers of German and in English with native speakers of English in order to better understand how the language one speaks influences the way words are perceived.
Specifically, the study sought to answer the following questions:
1. To what extent (if at all) does grammatical gender affect the perception of words?
2. How are conflicts between word meanings and grammatical gender settled?
3. How will diminutives be classified by German and English speakers?
In the first task, subjects were asked under time pressure to classify labeled pictures of objects according to their similarity with one of two possible reference pictures that represented biologically feminine concepts (queen, ballerina, or woman) or biologically masculine ones (king, giant, or man). Targets consisted of word-picture displays of masculine, feminine and neuter nouns as well as neuter diminutives in German for the German subjects and their equivalents in English for the English-speaking subjects. Directly following each response selection, subjects were asked to rate the similarity of the target and the chosen reference picture Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 29 on a scale of 1-5, with 5 being the most similar. The similarity rating task was included in order to evaluate subjects’ awareness of their gender-based associations and the strength of those associations, as well as to determine whether German speakers have stronger associations between the depicted objects and biological gender than do English speakers.
It was predicted that German speakers would be biased by their meta-knowledge of grammatical gender in pairing presented target words with biologically masculine or feminine pictures. Because of the more covert noun class system of English, native English speakers would not have such biases and would classify masculine and feminine nouns more unpredictably than German speakers. Furthermore, it was predicted that for diminutive targets derived from masculine and feminine root words, German speakers would be unable to classify these words as neuter (i.e., according to grammatical gender as determined by their affixes –chen and -lein), and thus would have three possibilities for classifying the nouns: 1) according to the grammatical gender of the word’s root, or 2) according to the apparent ‘masculinity’ or ‘femininity’ of the diminutives, or 3) arbitrarily with no apparent pattern. Responses of English speakers were expected to follow the pattern in 2) or 3).
The rating task should serve to indicate whether German speakers have stronger associations between grammatical gender of depicted nouns and biological gender than do English speakers. Thus, if there is a pure gender effect for German speakers we will see 100% of feminine words being paired with biologically feminine reference pictures, 100% of masculine words being paired with biologically masculine reference pictures, and neuter words will fall evenly in between. If we see purely an effect of culturally defined femininity with the diminutives, all diminutives should be perceived as feminine, and such a result would conflict with the conclusions of Boroditsky, et al. (in press).
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 30
Thirty-six subjects participated in the German part of the experiment and 37 participated in the English study. The German subjects, recruited with assistance from staff members from the University of Tübingen, consisted of 25 females and eleven males. All were native speakers of German and students at the University of Tübingen, and they ranged in age from 19 to 31 years with an average age of 21.2 years. Every subject had studied at least one foreign language with a gender system in addition to English. Subjects were paid €10 for their participation.
English-speaking subjects, ranging in age from 18 to 43 years with an average age of 21 years, were recruited from the undergraduate population at the Ohio State University. Many had studied a foreign language (usually Spanish) in high school, but most reported low proficiency in it and other foreign languages. Three of the English-speaking subjects grew up bilingual (English-dominant) and spoke English and one other language (German or Igbo), five reported fluency in a language they had learned in school (American Sign Language, German or Spanish).
None knew the goals of the study before they participated, and each participated in the study in exchange for course credit.
Stimuli were presented on a computer display programmed with E-prime (Psychology Software Tools, Inc.). Stimuli for each language consisted of labeled pictures for nouns that were selected according to gender categories of German. The labels for nouns did not include the definite article for German words (see Figure 7.1). Word labels for the English-language stimuli were derived from the English equivalents of the German words. Target stimuli included 29 pictures of grammatically feminine nouns, 19 pictures of diminutive nouns formed from Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 31 grammatically feminine nouns, 27 pictures of grammatically masculine non-diminutive words, 20 pictures of diminutive nouns formed from grammatically masculine roots, 14 pictures of grammatically neuter words, and 15 pictures of grammatically neuter words with diminutive labels. Of these, only 15 of each category were used in the data analysis. Pictures were hand drawn and were presented in black and white. All pictures used in the diminutive condition were identical to their non-diminutive counterparts; they differed only in their labels.
Subjects first saw a picture of a target object and were asked to pair it with one of the two reference pictures located to the left and right above it. For clarity, a text label of the target was written below the picture (in German for German speakers, in English for English speakers).
The paired reference pictures (man/ woman, king/ queen, giant/ ballerina) each represented biological gender. Left/right position of the reference pictures was counterbalanced across trials.
After subjects responded to each test slide by pressing a button on a response box, they saw a slide that prompted them to give a similarity rating from 1-5, with 1=most similar and 5=least similar, for the target word and the reference picture they had selected.
To discourage subjects from engaging in a conscious strategy in responding, the experiment was a speeded task. Subjects were instructed to respond as quickly as possible with their best guess even if the target and reference pictures seemed unrelated. Although subjects were asked to keep their responses under four seconds, the slide remained on the screen until a response was given. This was important because in Boroditsky’s experiment, there was the possibility that subjects had enough time to engage in a conscious strategy for completing the tasks (e.g., consciously producing feminine adjectives for feminine nouns presented to them).
The average response time for German-speaking subjects was 2443 milliseconds and the average response time for English-speaking subjects was 1036 milliseconds. There were no significant differences among the response times, so they will not be discussed further.
The 197 stimuli consisted of 133 test slides and 57 filler slides to discourage the adoption of strategies to respond, and seven control slides, which were included to ensure that the feminine and masculine reference pairs were perceived by subjects to be more similar to their respective gender categories than to the opposite gender. The seven controls consisted of reference pair members (giant/ballerina, man/woman, and king/queen) presented as targets to be matched with other members of the reference picture set, (e.g. ballerina was shown as a target to be matched with either king or queen). The fillers were chosen to discourage awareness of the
goal of the experiment. Fillers were broken down into the following types:
Breakdown of filler stimuli included for the German experiment.
Subjects were tested in their native language with all instructions and materials presented in German or English, respectively. After completing the experiment, subjects were asked to comment on what they thought was being tested in the experiment. Comments were recorded by the experimenter in field notes. Common answers given by German speakers were “the association between words and genders.” When asked to explain, they often cited that when they saw ‘kitchen’, they thought they should pick the feminine referent picture because women are often associated with cooking and cleaning. Several participants responded that they thought the experimenter was researching gender stereotypes. There were only a few who reported that they had no idea. Although subjects may have surmised that the experiment concerned biological gender because the reference pictures consistently provided a ‘masculine’ and ‘feminine’ version of the same category (e.g. ‘king’ and ‘queen’ are male and female versions of a monarch), with the exception of one subject, they did not seem to be aware of a manipulation involving the grammatical gender of the target object picture. The responses of one exceptional German subject who guessed exactly what was being tested were excluded from data analysis. The responses from the German participants stand in startling contrast to those of participants in the English portion of the experiment, where the vast majority of subjects answered that they had no Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 34 idea what was being tested. This suggests that German speakers may be more conscious of gender.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 35