37.3514, Software: Marcadores de Paula (ver. 0.9.1)

[The LINGUIST List]

LINGUIST List: Vol-37-3514. Mon 26 Oct 2026. ISSN: 1069 - 4875.

Subject: 37.3514, Software: Marcadores de Paula (ver. 0.9.1)

Moderator: Steven Moran (linguist at linguistlist.org)
Managing Editor: Valeriia Vyshnevetska
Team: Helen Aristar-Dry, Daniel Swanson
Jobs: jobs at linguistlist.org | Conferences: callconf at linguistlist.org | Pubs: pubs at linguistlist.org

Homepage: http://linguistlist.org

Editor for this issue: Daniel Swanson <daniel at linguistlist.org>

================================================================


Date: 25-Oct-2026
From: Emma Machado [emmachado at usal.es]
Subject: Marcadores de Paula (ver. 0.9.1)


Marcadores de Paula is a free and open-source desktop application for
transcribing and annotating recorded speech. It was developed for
research on discourse markers in spoken Spanish, where repetitions,
fillers and restarts are part of the data, so it leaves them as
transcribed and keeps the machine transcript next to the corrected
one.
The machine transcript is produced on the user's own computer, from
local audio or video files or from individual public videos downloaded
with yt-dlp. WhisperX transcribes these recordings with Whisper
large-v3-turbo by default, or large-v3, aligns each word to the audio
and uses pyannote to determine who speaks when (speaker diarization).
The machine transcript keeps Whisper's wording, with no prompt or text
normalisation added. Whisper may still leave out disfluencies or
replace words, so the transcript is checked in a review editor, where
each turn plays back with its audio and its words, speaker label and
boundaries can be corrected.
The corrected transcript is annotated on a separate layer with a
project codebook. Its 27 labels, partly based on the Val.Es.Co.
transcription conventions, and 21 discourse functions can all be
edited. A search for common Spanish marker forms and repeated adjacent
words proposes candidates, which have to be confirmed by listening.
The codebook and the search are designed for Spanish, while
transcription works in any language Whisper supports.
Transcripts export to TXT, DOCX, CSV, JSON, SRT, VTT, Praat TextGrid
and ELAN EAF, with a tier per speaker in TextGrid and EAF and the
annotations as optional further tiers. Annotations also export on
their own, including as concordance and count tables.
Processing time has been measured on one computer, a MacBook Pro with
an Apple M5 Pro and 48 GB of memory, using the CPU engine. On that
computer, a one-hour excerpt of a Spanish podcast with four speakers
took 42 minutes with large-v3-turbo and 53 with large-v3, alignment
and diarization included, in under 5 GB of memory. The Windows engine
for NVIDIA graphics cards has not been tested yet.
The application requires 64-bit Windows, or a Mac with Apple silicon
and macOS 26 or later, and its interface is in English. Diarization
needs a free Hugging Face account for the pyannote model. Because the
Windows installer is not code-signed and the Mac version is not
notarized, the README explains how to open the application the first
time.
Licence: GNU General Public License, version 3 or later
Downloads and documentation:
https://github.com/emmachados/marcadores-de-paula
Citation (all versions): https://doi.org/10.5281/zenodo.23482888

Linguistic Field(s): Discourse Analysis
                     Pragmatics
                     Text/Corpus Linguistics

Subject Language(s): Spanish (spa, stan1288)


------------------------------------------------------------------------------

********************** LINGUIST List Support ***********************
Please consider donating to the Linguist List, a U.S. 501(c)(3) not for profit organization:

https://www.paypal.com/donate/?hosted_button_id=87C2AXTVC4PP8

LINGUIST List is supported by the following publishers:

Australian Linguistics Society https://als.asn.au/Home

Bloomsbury Publishing http://www.bloomsbury.com/uk/

Cambridge University Press http://www.cambridge.org/linguistics

Cascadilla Press http://www.cascadilla.com/

De Gruyter Brill https://www.degruyterbrill.com/?changeLang=en

Edinburgh University Press http://www.edinburghuniversitypress.com

European Language Resources Association (ELRA) http://www.elra.info

John Benjamins http://www.benjamins.com/

Language Science Press http://langsci-press.org

MDPI Languages https://www.mdpi.com/journal/languages

MIT Press http://mitpress.mit.edu/

Multilingual Matters http://www.multilingual-matters.com/

Narr Francke Attempto Verlag GmbH + Co. KG http://www.narr.de/

Netherlands Graduate School of Linguistics / Landelijke (LOT) http://www.lotpublications.nl/

Peter Lang AG http://www.peterlang.com

SIL International Publications http://www.sil.org/resources/publications

University of Toronto Press https://utppublishing.com


----------------------------------------------------------
LINGUIST List: Vol-37-3514
----------------------------------------------------------