For Researchers
CQC is open to non-commercial research in linguistics and in related fields. You do not need a background in phonetics to use it. This page shows what the corpus supports and what to check before you apply.
What the corpus contains
Approved users receive transcripts, Praat TextGrids, speaker and session metadata, timestamps, and audio files. See the corpus documentation and the datasheet for details.
What you can study
The examples below are possible directions, not a guarantee that the current corpus can fully answer each question. Whether it fits depends on your research design and on the limitations described on the Corpus page.
| Field | Example questions | Keep in mind |
|---|
| Phonetics and sociophonetics | How do speakers differ in pitch, vowels, voice quality, or speaking rate, within and across groups? | A group difference does not mean every individual follows the same pattern. |
| Sociolinguistics and discourse | How do speakers present themselves, take positions, and use community labels such as 酷儿, 无性恋, or 跨性别? | Speech features cannot be interpreted apart from their context. |
| Gender and sexuality studies | How do speakers talk about identity, community, and personal experience in public videos? | The speakers do not represent all members of sexual and gender minority communities. |
| Sociology and anthropology | How do public narratives reflect community practices and self-presentation online? | A public video is not a full record of a person’s life or experience. |
| Speech technology | How well do speech recognition and other systems work across speakers and recording conditions? | Results must be reported in aggregate; classifying identity is prohibited. |
Different projects need different data. Some can start from transcripts and metadata; others need the time-aligned speech. Say in your application which data you need and why.
Researcher checklist
- Read the corpus documentation and the datasheet.
- Read the Data Use Agreement, especially the prohibited uses.
- Obtain ethics approval from your institution if it is required.
- Decide whether your project really needs speaker-level identity information.
- Report sample sizes, imbalance, and limitations.
- Refer to speakers only by CQC speaker IDs, and cite the CQC release you used.
How to get access
- Read the documentationCorpus page, datasheet, and agreement.
- Apply by emailSend the application template on the Data Access page.
- ReviewWe review applications on a rolling basis and may ask follow-up questions.
- SignThe PI and an institutional official sign the agreement.
- DownloadYou receive a personal download link.
Go to Data Access
Related corpora
CQC is part of a small but growing set of resources on gender- and sexual-minority speech. The list below is assorted rather than complete, and each resource has its own design, access rules, and limitations. Check the original publication before you use or cite it.
Speech corpora of gender- and sexual-minority speakers
| Resource | Language and speakers | Access |
|---|
| Queer Waves | German. About 335 hours from 407 self-identified LGBTQIA+ speakers, from podcasts and YouTube. | Summary statistics public; speaker metadata and download scripts on request, for research only. |
| Gender Spectrum Speech Corpus | English and French. Read speech and interviews with speakers across the gender spectrum, including non-binary and trans speakers. | Version 2.1 distributed through ORTOLANG; check the current terms. |
| MAGES Corpus | American English (Mid-Atlantic United States). 14 transgender and non-binary speakers, each reading 399 sentences. | After an appointment with the lead researcher to discuss the intended use. |
| A Palette of Transmasculine Voices | American English. 20 masculine-identifying speakers, including trans men, cis men, and transmasculine non-binary people. Read passages and sentences, with Praat TextGrids for the sentences written for the project. | Open access; intended for clinical and teaching use. |
Spoken corpora for discourse and lexical research
| Resource | Language and speakers | Access |
|---|
| Corpus oral de la comunidad de habla LGTBI | Spanish (Spain). 5 informal conversations, 20 semi-structured interviews, and social-media texts from members of the LGTBI speech community in Valencia, Madrid, and Barcelona (Navarro Carrascosa 2023). | Transcripts and speaker metadata published in a book (Editorial Universidad de Alcalá); audio not included. |
| CoLaGe | Spanish, from Valencia (Spain) and Guadalajara (Mexico). Sociolinguistic interviews and role plays, designed to study gender, sexuality, and language use; not limited to queer speakers. | Language Bank of Finland: text versions for academic users; audio versions on application. |
General speech datasets with some gender-diverse labels
These datasets offer gender options beyond the binary, but very few speakers use them. An audit of six English speech datasets found that 0–1.4% of speakers identified outside the gender binary (Sheppard et al.).
| Resource | Gender information | Gender-diverse speakers |
|---|
| Casual Conversations v2 | Five categories: cis man, cis woman, trans man, trans woman, non-binary. | About 1.4% of speakers. |
| Edinburgh International Accents of English (EdAcc) | Free-text self-description. | About 0.8% of speakers. |
| Mozilla Common Voice | Options include transgender, non-binary, and intersex (since 2024). | Very few; published counts for English differ (0 or 11 of about 95,000 contributors). |
| German Plapper Corpus | Gender (diverse, female, male, no answer), and sexual orientation on the Kinsey scale. | 8 of the 1,522 speakers analysed in 2024 chose “diverse”. |
Collections of Chinese queer materials
| Resource | Contents | Access |
|---|
| The Leiden Chinese Queer Collection | Primary sources and scholarship on the Chinese queer experience across the Chinese-speaking world and its diasporas: literary works, films, monographs, journal publications, and magazines. Not a speech corpus. | Held by Leiden University Libraries; see the project page. |
We know of no other public corpus of Mandarin or other Chinese speech from gender- and sexual-minority speakers. Acoustic studies such as Geng and Gu (2022) on gay and heterosexual men and Huang et al. (2024) on trans women are, as far as we know, not released as corpora.
Know of a resource we missed? Write to chinesequeercorpus@outlook.com.
List compiled October 2026.