I'd like to know the difference between the way lexicographists wrote dictionaries in the past, and the way they could procede now, in our modern times.

Note: To compare, I'm talking about lexicographists who work alone on a project, not a team. And a project with a big amount of words, not a small glossary, because the comparison couldn't apply.

For instance, Émile Littré wrote a French dictionary alone, and María Moliner wrote her Spanish dictionary "Diccionario del uso del español".

How did they do that? Did they wrote all the words and expression they heard? Did they use a previous dictionary to improve it?
How did they write their notes, to sort in alphabetical order the words, each time they needed to add a new word between two letters?

How to do that now? What changed in the way they did it, and the way we can do it now? How to do now?

It seems a lot of question, but it's only to guide the answers. My question is what are the differences between the way they did before, and the way a single lexicographist could procede now. What is his modern method?
Isn't the fact to make a dictionary really faster nowadays? Why?

There's already a similar question here, about the method: How are dictionaries produced

Explaining how dictionaries are made, but it doesn't ask the question of the comparison between old and new methods. (I also mean a method not with a software making analysis of everything and doing the work, but a more artisanal method, while using technology, but as someone who is not particularly equipped in professional softwares, as a lambda person)

  • 2
    You will get very different answers, depending on whether you are restricting this to unabridged dictionaries of major literary languages, vs. dictionaries (of perhaps 5,000 words) for any available and possibly unwritten language.
    – user6726
    Commented Nov 11, 2019 at 18:55
  • 2
    They used to have drawers with cards in alphabetic order, you can easily insert new cards between the existing ones, you can easily rearrange the cards, etc. First they collected the materials, then they rewrote that from cards into a book. And now they use computers.
    – Yellow Sky
    Commented Nov 11, 2019 at 23:40
  • The drawers with cards is very interesting, and would worth to be an answer. The difference between kinds of dictionary, by user67, too.
    – Quidam
    Commented Nov 11, 2019 at 23:48
  • How did they collect the material?
    – Quidam
    Commented Nov 11, 2019 at 23:49
  • Why downvoted???
    – Quidam
    Commented Nov 12, 2019 at 9:21

2 Answers 2


The impact of technology for undescribed language lexicography is that it expands the set of languages for which dictionaries exist. This is due to two factors. First, computer technology makes it much easier to gather and organize the underlying data. Second, said technology makes it possible to disseminate materials which could not have been disseminated in the pre-technological era.

The no-tech method of assembling data primarily involves elicitation onto paper, and post-hoc extraction of data into lexical slips / cards. In case there is some literature on the languages, you can pile books and newspapers up in an office and at some point transcribe examples onto cards. This is labor intensive, and requires that you know how the information on the cards should be structured (which you usually do not know until you have been working on the project and discover that you have to re-do all of your old cards). When you get to the point that you think you're done, you start typing entries up (possibly also handwriting special characters), according to the publisher's format requirements. Finding a publisher can be difficult given the expense of producing print books from typescript.

Computer technology makes it vastly easier to pull together the primary data: your elicited data is already in a computerized form; you can also call on electronic publications on the language, including scanned and OCR'd print works. Computerized data is much easier to manipulate compared to material on cards or written in notebooks – it's really hard to extract all instances of "chicken" from a paper corpus, and takes a minute with a well-structured electronic corpus. It is easier to add information incrementally and to re-conceptualize data-organization with digital data, compared to typed or handwritten information on paper.

Dissemination is easier because (1) you can do the formatting yourself, reducing costs (2) you can produce a PDF yourself and (3) online free dissemination is quite easy – no publication subvention is required. It is also potentially vastly superior in terms of informational content, since sound files can be linked to entries in an online dictionary. However, this feature still awaits certain technological improvements (good-looking output, embedded sound files, and manageable file sizes). The comparative ease with which a dictionary can be produced means that a dictionary becomes less of a life and resource commitment.


Part of making a dictionary is getting it to look right on the page. Once upon a time, I guess, that would have done by a professional printer, but these days our local press wants camera-ready copy. I've formatted several dictionaries, so I know something about this end of it. I gather from your question that the actual printing does not particularly interest you, so I'll keep this short.

I wrote a bunch of Lex (Flex) pattern matching programs to read through the substantive part of a completed computerized dictionary and spit out TeX code which could be sent to a printing device understanding TeX. That prints out the dictionary, entry by entry, line by line, column by column, page by page, handling special characters, indentation, justification, line breaking, bold face, and italics.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.