Analysis
Character Count
Since the entirety of our corpus draws from written (in this case, typed) data, we chose to include character count as a potential factor in determining why a French speaker would choose to employ an anglicism over alternative lexical variants. We decided that character count, for our purposes, is rather comparable to syllable count in spoken speech and that functionally-speaking, a speaker would theoretically want to choose words that take the least amount of time to type—efficiency being a primary motivator in discourse decisions.
Here, we see that over half of the anglicisms we found fall under the medial category of “between 6 and 9 characters,” while roughly one-third consist of between 1 and 5. The smallest group, comprising 12.9% of the data, contain 10 or more characters in total.
Adopted English words, then, in general tend to be rather short, and based on the measly number of anglicisms within the 10+ category, it is not unreasonable to assume that an inversely correlational and possibly causational relationship between character count and likelihood to be borrowed does, in fact, exist. Overall, there is an 87.1% chance that the borrowed word will fall into one of the first two categories, whereas there is only a 12.9% chance that the anglicism will consist of more than 9 characters.
There is, of course, the issue that the majority of words in both English and French do not consist of more than 9 characters. Even in the previous sentence, for example, the only word that consists of 10+ characters is “characters” itself. To know for certain if an individual’s choice is at least partially influenced by character count, then, it would become necessary to compare each anglicism to the character count of its suggested replacement. For now, missing this data, we can offer the idea that such a relationship may exist, but we do not have adequate data to say for certain whether or not it does. Such a relationship, among other factors, could work to explain a French BuzzFeed user’s decision to employ “hashtag” (7 characters) over the Académie’s “mot-dièse” (9 characters), among others.
Syllable Count
Like character count, we included syllable count in our analyses in an attempt to relate this written data back to the spoken, community object that sociolinguists traditionally prefer to study. Syllable count relates back to the linguistic efficiency principle—if it can be explained in less time, using fewer words, why would we want to make things more difficult for ourselves? Theoretically-speaking, then, a speaker should exhibit a preference for words with fewer syllables, and when deciding between anglicisms and Académie suggestions, speakers, overall, should exhibit a tendency to choose the option with fewer syllables.
Here, we see that words consisting of 4 or more syllables seem to be highly disfavored, and excluding the 1 syllable category, there is a notable inverse correlation between word-length and an anglicism’s likelihood to be adopted. Overall, based on our data, there is a 94% chance that an anglicism will consist of 3 syllables or fewer, meanwhile there is only a 6% chance that an anglicism will be longer than 3 syllables.
As is true for the character count analysis, it is possible that these findings falsely identify a relationship between these two factors and that the true motivator behind these numbers is, in reality, simply a lack of lengthy English words available for borrowing. To test the relevance of this correlation, then, it would become necessary to extend the current study and directly compare the syllable count of each anglicism to that of its corresponding Académie suggestion.
Parts of Speech
The overwhelming majority of anglicisms were nouns, which make up 74.8% of our results. The next largest categories were adjectives, many of which were derivatives of these nouns (i.e. tweeter), at 14.6%, and then verbs at 5.3%. Exclamations, abbreviations, interjections and adverbs together made up the remaining 5.9%.
These results, in general, are not very surprising, as the syntactic category of “noun” tends to be the most open to neologisms, and therefore, contains the largest pool of lexical items from which other languages may then borrow tokens. Additionally, many of these anglicisms are products and inventions, which are typically more difficult to translate than a word like “tree,” for example, for which French already has a direct correspondent in the form of arbre. Finding French equivalents for words like “Sloppy Joe” and “selfie-stick” would unquestionably be significantly more difficult, and the resulting translations would likely fail to capture the words’ full meanings.
Semantic Analysis
After finding all of the anglicisms within our corpus, we divided them into semantic categories. Overall, the largest concentration of tokens seems to fall under “Food” at 17.6%, followed by “Media” at 10.7%, and words associated with “Persona/Identity” at 9.4%. A possible explanation for these numbers could be the lack of French tokens available to use instead. English words for animals, for instance, often have exact semantic matches in French (i.e. dog/chien), whereas there are no French equivalents for words like “donut” and “mash-up.” One could theoretically choose the French word for pastry, “pâte,” over donut, for example, but then listeners would have a harder time determining exactly which kind of pastry the speaker means. Choosing to use the anglicism in these situations, then, is a functionally logical choice, as the anglicism better captures the meaning the speaker wishes to evoke and thus, increases conversational clarity. Another possible explanation, which could potentially work to explain the usage of anglicisms like “ticket” (French eqv: ”billet”), is that there may actually be prestige associated with using anglicisms. According to the Pew Research Center’s ongoing “Global Attitudes” project, for example, 75% of French citizens have a “favorable” view of the United States. It is possible, then, (assuming this number is truly reflective of modern French society), that the choice to use an anglicism over its French equivalent may actually, in some cases, be a stylistic one. It may be that a positive view of American culture has led to a positive view of the English language and that, contrary to what the Académie may wish, French speakers now show an affinity for all things American, including our language.
Token-Type Analysis
To find the number of anglicisms compared to the total word count, we completed a token type analysis, and arrived at 1.45%. Compared to the other studies, such as the Bogaards 2008 study of borrowings in French, where researchers found a type-token percentage of 2.6%, our number reflects a near 50% decrease. As one of the aforementioned downsides of our BuzzFeed corpus is the disproportionately large focus on American topics, we would have actually expected to see a number larger than 2.6%, not lower. Additionally, many of the anglicisms we found in our data (i.e. selfie stick, Instagram, #tbt) did not even enter the English language until after 2008, meaning the amount of anglicisms present within French has to have grown somewhat since 2008, even if only negligibly.
Unfortunately, this means that it is very likely that the 1.45% figure is incorrect, as this would imply that there has been drastic decrease in anglicisms since 2008, while we have data from post-2008 suggesting that there has to at least have been some growth.
