A very late reply from me here! There’s so much to say, but I’ll speak first of the paper I found most interesting: How a Dictionary Became an Archive: Community Language Reclamation Using the Mukurtu Content Management System.
This is about Mukurtu, which is a “free, mobile, and open-source platform built with Indigenous communities to manage and share digital cultural heritage”, and its deployments for various indigenous communities. Mukurtu is a content management system (this is the same genre that Drupal belongs to) which specializes in serving indigenous communities.
The most interesting thing about this system is, I think, how widely its scope its conceived. While a Mukurtu deployment is often called a “dictionary”, as the authors explain, it is in fact also used for plenty of non-linguistic content. For example, while there still might be entries where there’s a headword, some lexical information, and a sense inventory, there might also be photo and video data portraying cultural artifacts of interest such as textiles. Language, then, is just another kind of cultural heritage next to other kinds on this platform.
The word “dictionary” is used even as Mukurtu’s creators acknowledge that this it does not conform to the traditional conception of such a product. The justification for this is, for one thing, that some indigenous communities prefer to use this word:
As Wynne recalls, when asked what they would like this set of language tools to be called, most community users suggested that it be labeled simply Dictionary . This term was familiar, indicated a collection of language materials, and helped provide contrast with Digital Heritage Items (DH Items, explained in the following section) that were not explicitly linguistic in nature that had also been uploaded to the platform.
In other words, “dictionary” is an already-known word that communicates much of the intent behind Mukurtu (a place for curated cultural heritage), and it is simpler to coerce this mostly-right word into an expanded meaning which also covers non-linguistic artifacts instead of to try to coin a new one.
But even if we consider just the narrow conception of what a dictionary is, Mukurtu also intentionally breaks with traditional approaches to lexicography. Mukurtu has fine-grained authorization features which allow communities to decide who has read/edit permissions to any kind of content, which includes dictionary (in the narrow sense) entries. In the case of Pala, the community began by importing a dictionary made by Jane Hill, and used it as a basis for a continually-updated dictionary hosted within Mukurtu. This decentralizes the dictionary: instead of having all changes flow through a narrow bottleneck (e.g. the single documentary linguist who works on a given language), community members are empowered to edit the dictionary as they see fit:
Relatedly, the way that the Mukurtu Dictionary is designed does not center the types of expertise required by traditional lexicography. An “expert” in this context […] knows exactly which knowledge can and should be available to dictionary users. These issues arose during a discussion with those involved with the Pala Dictionary. Users had noticed that some of the materials in Jane Hill’s A Grammar of Cupeño (1966) didn’t capture the increased use of schwa [e], a language change that has occurred in the last fifty years (one can assume, due to increased contact with English). So, as they put it, they “corrected it,” including the Hill words in the Alternative Spelling fields, establishing themselves as experts on local pronunciation. Alicia, a staff member who has contributed to the Pala language program for years, said:
Okay, well, this is what Jane Hill did. She’s a queen. Boom. This is her work that she did. It’s in this book. She made mistakes. We all make mistakes. Some of these things do need to be fixed […] But yeah, the ability to add alternate spellings is a great tool. I wouldn’t call it a secondary language, a dialect or anything. It’s just alternate spelling (Interview, July 2023).
This is an approach with clear benefits, though one wonders what happens when there is not consensus in the community about how to handle any particular linguistic phenomenon: what if some in the community had felt it important to not modify the spellings for whatever reason? As far as I recall, how conflict like this is resolved is not really discussed in the article, but perhaps the answer is that it would simply be upon the community to resolve this conflict just like any other conflict that might arise over how to manage cultural artifacts.
Beyond conflicts like this, I also wonder what a dictionary edited in this manner looks like after 20 years. If “non-expert” (in the very narrow sense that they do not have formal linguistic/lexicographic training) users exclusively curate a linguistic dictionary for 20 or 40 years, will the product end up being something that the community will find useful in the same way that they found the Jane Hill dictionary useful? I should sincerely hope so, but it is not clear to me what the qualities of a dictionary maintained in this way would be, and it seems imaginable to me that a community might adopt permissive editing policies and then, some years later, come to wish that there had been tighter editorial control.
Mukurtu deployments are relational in the sense that any content is linkable to any other content, and it is therefore very easy to e.g. link from a sense under a headword to a photo or video illustrating the sense. You might, for example, link from a word for a specific textile to a photo of it or a video of it being woven. On this, the authors write:
The importance of the Relations feature cannot be overstated. It is a concrete critique of dictionaries as simple collections of words and pushes against the dominant Western language ideology that holds that language is a separate domain, an object that can be studied synchronically, without consideration for history or context.
In other words, the classical dictionary is deficient in the sense that the medium inherently conceives of language (or words) as decontextualizable. If the suggestion is that using a classical dictionary perpetuates the “dominant Western language ideology”, or even that if one does not subscribe to the dominant Western language ideology, then one ought not to publish a classical dictionary, then I think this is too strong: there are of course plenty of lexicographers who would vociferously subscribe to the views espoused here and nonetheless publish a dictionary for the simple reason that (classical) dictionaries, for all their flaws, are practical, and before the digital age, were also (I think it is fair to claim) the only way to systematically disseminate a record of the lexical inventory of a language. (To see this, consider how big a print dictionary would need to be if it aimed to be a record of not just linguistic facts but also every potentially relevant cultural one for each word.) So, even if we grant that the “dominant Western language ideology” is to be criticized, I do not think it necessarily follows that one should not be interested in a traditional dictionary. And perhaps the authors would readily agree with me here—it’s not clear to me. But this passage could plausibly be taken as an argument against using classical dictionaries at all.
Later in the conclusion, this matter is picked up again (emphasis in boldface mine):
The need to devote so much time to non-lexicographic features of the platform within a paper about Indigenous language dictionaries illustrates our most important takeaway: that language and culture are inextricable and deserve to be treated as such. Unlike a traditional dictionary, an archive—like a community—is never truly “finished,” but only becomes richer as it grows and changes. When a dictionary is more like an archive, one can avoid the view of a language as a bounded “thing,” able to be contained in a discrete, alphabetized volume. Mukurtu’s structure engenders a collaborative social project, one that pulls people together to make their own decisions about meaning, context, translation, and approaches to cultural heritage management.
Here again I think that if the suggestion is that a classical dictionary entraps us in a problematic language ideology, this is seems overly deterministic. McLuhan was right that the medium is the message, but this does not rob deprive producer or perceiver of all agency in how some content is to be understood in a certain medium.
This is not to detract, of course, from what is in my mind the distinctive accomplishment of Mukurtu: providing a technical means for centering community members in the lexicographic process. And it is also right that the form of the traditional dictionary embodies certain language ideologies, and so it is good to have alternatives, especially in cases where communities feel strongly about the medium that is used.