Edits to Preliminary Kpelegbe Dictionary

The original version of the Kpelegbe dictionary was a set of hand-written fieldnotes from the early 90s, meant to provide me with a tool for my syntactic analyses at the time. It was arranged by tonal category (e.g., L.LH).  During this last academic year (2025-2026), I wrote up those notes into a word file, and gave it to Zach to translate into a FLEx project, which is much easier to work with in creating dictionaries. That initial process is outlined here:

First Steps: Writing the Kpelegbe (Ewe) Dictionary

Since Zach created the first Flex version of the Kpelegbe dictionary on March 20, 2026, I have gone through four versions (today is May 8, 2026). By version, I mean a printed-out form of the dictionary that I can read for errors. Although the dictionary is not that big as of now (around 700 words), making all these changes early on sets me up for a smoother process down the road. 

I went through the same process of lengthy editing with the Sasi dictionary (also created with FLEx). That was a learning experience for me that has helped me immensely in facing the challenges of the Kpelegbe dictionary. In fact, had I not written the Sasi dictionary, I doubt I would have had the courage to start in on the Kpelegbe dictionary. See the following link for some background on the Sasi dictionary:

Writing the Sasi Dictionary: Some Personal Recollections

Major Themes

1. To make the dictionary as user-friendly as possible for the Kpele community.

2. To make the transcriptions as much like standard Ewe as possible, while retaining Kpelegbe pronunciations. My model dictionaries for this process are Westermann (Ewe-English) and Rongier (Ewe-French), two well-known dictionaries of standard Ewe.

3. To make the entries consistent, adopting uniform formatting conventions throughout the dictionary (following as closely as possible from points 1 and 2 above).

Edits Made

Created a cover page with the name of the dictionary.

Changing all schwa transcriptions to ‘e’ to make transcriptions look more like standard Ewe. This change was done in the spirit of making the dictionary more user-friendly.

Added notes in the dictionary where the symbol ‘e’ is pronounced like [e] and not like schwa (there are only a small number of such words). A typical such note is ‘Second vowel is [e].’

Fixed the symbol for the bilabial fricative, which I had been writing with an italic ‘f’ (which does not have the right Unicode).

Reformatted pronoun entries. I had written the definition with symbols like 3SG, which are not user-friendly for non-linguists. So I changed the definition of pronouns to the English word (and French word) (e.g., ‘you’, ‘me’, etc.).

Capitalized all proper names in dictionary, including place names.

Made sure the section headings of each section contain both an upper and lower case letter (some were left out of the first version). 

Corrected Kpelegbe transcriptions and tonal transcriptions. There were only a handful of errors. 

Added missing ‘to’ before verbs in English glosses. There were only a handful of instances.

Got rid of slashes “/” to show alternatives (e.g., ‘il/elle’ becomes ‘il, elle’). This change was just for uniformity in the lexical entries.

Made sure example sentences vary evenly between masculine and feminine translations for third person singular sentences. I became aware of this issue in writing the Sasi dictionary.

Set up an ‘alt.’ field for alternative forms. Very useful.

Deleted duplicate entries (from the initial written version). There were only a handful of duplicate entries.

Merged senses into one lexical entry (using FLEx merge). Senses are defined as phonologically identical words (including identical tone). There were only a handful of instances of merging that needed to be done.

Made the format for proper names uniform throughout. Kpelegbe (Ewe) names often have an explanation (e.g., ‘name of male born Friday’). The explanation is not the definition now, rather it is contained in a note.

Used ChatGPT to correct French errors. The errors were mostly errors in accents, but there were also some suggestions about glosses. There were many such errors, so ChatGPT was very useful. It saved me a lot of time.