«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 ...»
8.0 German Results Responses to the German picture-choice classification task are presented in Table 8.1.
Responses to the German picture-choice classification task (presented in proportions) A 3 (gender) x 2 (size: diminutive vs. non-diminutive) repeated-measures analysis of variance was conducted on the German picture-choice data using the SuperANOVA program, with subjects as a random variable. The data were arcsine transformed before being submitted to ANOVA. Significant main effects of gender (F(35,1)=3.94, p.05), and size (F(35,1)=17.63, p.001), were observed, as well as a significant interaction between gender and size (F(35,1)=6.18, p.01). The main effect of gender, shown in Figure 8.2, indicates that Germanspeaking subjects were more likely to match masculine target stimuli to masculine reference pictures than they were to match feminine or neuter target stimuli to masculine reference pictures.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 36
Proportion of masculine, neuter, and feminine nouns associated with masculine referents in the German forced-choice task.
Figure 8.3 shows the main effect of the diminutive suffix.
Subjects were more likely to associate diminutive nouns with feminine reference pictures than they were to associate nondiminutive nouns with feminine reference pictures, which indicates a ‘demasculating’ or ‘feminizing’ effect for diminutives, regardless of the grammatical gender assigned to the associated root noun.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 37
Proportion of diminutive and non-diminutive nouns classified as masculine in the forced-choice task.
Cell means for the interaction between gender and size are shown in Table 8.4.
Means for the interaction gender x diminutive-non-diminutive.
The interaction between size and gender is depicted in Figure 8.5, and suggests that the ‘feminizing’ effect for the diminutive nouns was greatest for the nouns formed from masculine roots. Planned comparisons confirmed that the difference between means for responses to masculine non-diminutive and diminutive stimuli was reliable, p.001, as was the difference between responses to feminine non-diminutive and diminutive stimuli, p.05. Thus, for nonSex and Size: The Influence of Grammatical Gender on Object Perception in English and German 38 diminutive forms, grammatical gender influenced the German-speaking subjects’ perceptions of words in a predictable way, but this effect did not hold for diminutives formed from masculine root words. Instead of using the grammatical gender as a basis for their decisions as they had done for the non-diminutive words, subjects appeared to have been influenced by a ‘smallness’ or ‘feminizing’ criterion with regard to the diminutives in general, and to those formed from masculine nouns in particular. However, it should be noted that the German diminutive suffixes –chen and –lein cause the resultant diminutive to be neuter. Since there was no neuter referent picture option, subjects were forced to classify the diminutives as either more similar to the masculine or feminine referent picture. In addition, though in the neuter and feminine conditions, subjects chose both masculine and feminine reference pictures at a comparable rate, resulting in performances that were close to chance, there was a numerical effect of grammatical gender in the predicted direction. Figure 8.5 is graphed on a proportion feminine scale in order to point out the “feminizing” effects of diminutives.
Average responses to the similarity rating task, in which subjects rated the degree of similarity of the stimuli to the reference pictures, are shown in Figure 8.6.
Average responses to the similarity rating task for masculine, feminine, and neuter nouns and for diminutives from masculine, feminine, and neuter roots. 1=not similar, 5=very similar.
As can be seen in Figure 8.6, subjects used the lowest end of the scale for their ratings, indicating generally that they did not find the target stimuli very similar to the reference pictures.
There were no significant differences in similarity ratings for masculine, feminine and neuter
target stimuli (masculine non-diminutive: 1.79, neuter non-diminutive: 1.73, feminine nondiminutive: 1.96, masculine diminutive: 1.88, neuter diminutive: 1.77, feminine diminutive:
1.93). The numeric pattern in the responses suggests that subjects found the feminine-gendered nouns and feminine-based diminutives slightly more similar to their reference pictures than the masculine or the neuter, and the neuter nouns and their diminutives the least similar to the reference pictures. The somewhat lower ratings for the neuter nouns could be expected if subjects used grammatical, gender-based criteria to make their judgments, since the reference pictures did not include a neuter option.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 40 These results support earlier claims that grammatical gender itself may influence the associations a speaker has with any given object and because the grammatical gender of words varies from one language to another, the perceptions of speakers may also vary along these lines.
However, the findings of this study also suggest that a speaker’s ideas about a word are colored primarily by cultural experience. That is, the semantic associations of a word, those that reflect cultural experience, can override grammatical associations in the classification of words in the minds of a speaker. These results further show that there is another type of linguistic marking— one that feminizes. It is apparent from this study that the perceived femininity of a word increases if the object is diminutized. In the case of diminutives, the grammatical gender is less important than the semantic and cultural associations tied to the meaning of the word. This evidence points to the possibility that speakers are indexing words largely by cultural and semantic criteria and the grammatical associations affect, but do not determine the perception of the word. It is for cultural reasons that in this experiment, when the neuter grammatical category was not an option, the diminutive stimuli were more often selected to be ‘more similar’ to the feminine reference pictures than to the gender of the root nouns.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 41
9.0 English Results As Figure 9.1 indicates, the English-speaking subjects chose the masculine reference picture 57% of the time, and there was a slight decrease in the percentage of ‘masculine’ responses to target words in the forced-choice task for diminutives, which suggests a similar, though less robust, feminizing effect for diminutives in English as compared to the German data.
English results. Proportion of non-diminutive and diminutive nouns classified as masculine in the forced-choice task.
As the German and English word lists differed slightly in the two experiments because of the difficulty of finding matched pairs of diminutives in English for the German target stimuli8, a subsequent comparison was made using only the words that had appeared in the word lists of both the English and German experiments in order to control for possible confounds due to This problem can be attributed to the high degree of productivity of the –chen and –lein suffixes in German.
English diminutive suffixes, -y, -ie, -et and –sie are far less productive than their German equivalents.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 42 inequalities in the number of inherently gendered nouns on either list. Results of the matched comparisons for English and for German nouns are shown respectively in Figures 9.2 and 9.3.
Proportion of masculine, feminine, and neuter non-diminutive and diminutive nouns associated with masculine reference pictures in the German forced-choice task A comparison of these figures reveals the extent to which German subjects were biased by grammatical gender. The masculine non-diminutive words in German were more often associated with the masculine reference pictures than the feminine non-diminutive words.
German neuter non-diminutives fall in the middle as predicted. Whereas the English words (which would be masculine in German) were also more often judged to be similar to the masculine reference pictures than either the neuter or feminine, there is a standard error of.056, as compared to German’s.045. English words which would be feminine in German also seem to follow the same pattern, i.e., they were less likely to be associated with masculine reference pictures, however the standard error in these responses is much higher,.292 as compared to.082 for the German data. The higher standard error in English responses indicates that the responses Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 44 of English-speaking subjects were less consistent when compared to the responses of Germanspeaking subjects.
As noted earlier, the diminutive suffixes in German are much more productive than those of English. Where German has two options for diminutives, English has at least four. One possible explanation for the larger standard error results for English diminutives is that the forms varied so much from word to word unlike in German where –chen and –lein predictably occur.
Note that although both the German and English speakers found the set of target objects in the masculine group to be more similar to the masculine reference picture, it cannot be said that this factor accounts for the pattern of results in the German experiment. In the German experiment, forced-choice results show the most masculine responses for masculine, the fewest for feminine, and a middle ground for neuter. In contrast, the English data show more masculine responses for feminine target pictures than for neuters.
Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 45
10.0 Discussion Although German responses were in line with the grammatical gender of the target stimuli, the data revealed a clear bias of subjects to classify stimuli as masculine. The highest proportion of feminine responses in the forced-choice task in any of the three non-diminutive categories was.50, or roughly at chance for the feminine word group. This could be a result of the dominance of masculine nouns in the language as demonstrated in Mills (1984). In her study, Mills found not only that there were more masculine gendered referents in the stories she analyzed, but also in the experimental section that grammatically feminine toys were given masculine names more often than grammatical masculine toys were given feminine names.
Furthermore, grammatically neuter toys were most often given masculine names. The finding of a bias toward the masculine is also reminiscent of Steinmetz’s hierarchical ordering of German gender, in which masculine is the default, unmarked gender.
10.1 Limitations and directions for future research As stated earlier, diminutives were thought to have a correlation with femininity and the results of this experimental study add credence to this assertion. It is unclear from the present results why that is so, i.e., whether the correlation between diminutives and femininity is based on the associations speakers have with smaller objects being more feminine or whether it is perhaps based on the correspondence between diminutives and the people who use them most, women. This possibility needs to be explored in further research.
One limitation of this study lies in the meanings of the diminutives themselves. There are three translations for the word ‘diminutive’ in German. The first, Verkleinerungsform, comes Sex and Size: The Influence of Grammatical Gender on Object Perception in English and German 46 from the verb verkleineren, which means ‘to make smaller’ thus it denotes smallness. A second sense of the word ‘diminutive’ in German is Verniedlichungsform, which comes from the verb verniedlichen, which means „to trivialize.“ The third, Koseform, is the ‘affectionate form.’ It is difficult to know which form, Verkleinerungsform, Verniedlichungsform, or Koseform, the German speakers were thinking about when they were responding to the stimuli.
The diminutives used in the experiment are used primarily in speech, and thus normally used in a context that could help sort out which form is being used. It was in an unnatural setting that each word was seen and evaluated by German subjects. If a German subject saw the isolated word Entchen, ‘duckie’, there is no sentence context, no tone of voice and no gesture that would normally accompany such a word to help the speaker know if the noun is being minimized in size or minimized in importance. English speakers may not be aware, but they too make such
distinctions. To exemplify these differences in meaning I offer the following examples:
1. Small: Grandma says to the child “put the mousie down” and the girl releases her thumb and forefinger and sets the week-old mouse free.
2. Trivialized: A boyfriend says to his girlfriend who has climbed onto the kitchen table, “You aren’t going to let this little mousie scare you, are you?”
3. Familiar/Affectionate: When it’s dinner time, Jared says to his pet rodent, “Does mousie want some food?” Because there are so many semantic connotations a diminutive could have based on the context in which it is spoken, it is difficult to say with without a bit of doubt, which type of diminutive speakers were responding to.