Abstract

Genericide is the process by which a trademarked term is used generically by the public and ultimately loses its legal trademark protections. The linguistic methods that courts have used to determine whether a given term is in the process of or has undergone genericide have historically relied on dictionaries. However, there has a been recent push to use corpus linguistics as a tool to aid in that determination for not only trademarks but word meaning in general (Hoopes, 2019; Lee & Mouritsen, 2018). In addition to corpus data, I argue that the use of experimental data via a linguistic questionnaire can support, validate, and clarify corpus findings and can be an additional means to aid in the determination of the semantic status of trademarked terms. Corpora comprised of texts from the social media website Reddit were created and concordance lines exhibiting uses of 24 terms (10 generic and 14 trademarked) were judged based on their semantic senses as interpreted by two raters. These concordance lines were compared to the responses of a linguistic questionnaire asking participants how they used those 24 terms. Results show that the questionnaire responses are comparable to and validate many of the results of the judging of the Reddit corpora. The questionnaire data provided clarity on use of terms deemed ambiguous by previous research. I assert that the use of questionnaire data is a useful option in researching the genericide phenomenon either in conjunction with corpus data or independently. Both methods are considered helpful for courtrooms and businesses in investigating genericide, but based on the findings of this thesis I advocate that neither method can determine genericide alone but should be only considered as aids to work in conjunction with other evidence and data.

Degree

MA

College and Department

Linguistics and English Language

Rights

https://lib.byu.edu/about/copyright/

Date Submitted

2021-12-08

Document Type

Thesis

Handle

http://hdl.lib.byu.edu/1877/etd11969

Keywords

genericide, trademark, corpus, questionnaire

Language

english

Share

COinS