Diary of a Digital Apprentice (2): First, a Unix tutorial

(Here for the blog-series kick-off post).

We’re playing catch-up a little, and these are things I did in the tail end of 2017.

It’s been a long time since I’ve done anything with Unix. About 10 years, actually, and my unix experience was limited to running Ubuntu at the time and being forced to troubleshoot a lot of things mainly by googling answers. That was frustrating and satisfying at the same time. A memorable highlight was the time that my system switched to Ancient Greek at some fundamental level so that I couldn’t log in because it would only input Greek characters and it was not as simple as ‘change keyboard’.

Anyway, Jedi master Tauber decided I should learn to manipulate text files in Unix and set me the following tasks. You can see them over here:

This is what I call “hunt”-learning. The teacher isn’t pushing, and the learner isn’t actively trying to pull things from the teacher, rather the teacher is setting up tasks which the learner must then go and problem-solve. I think there’s a lot to be said for such a method, and it works particularly well for something like this.

Also, by the end of 7 tasks, I had not only an appreciation for how to do these things, but a sense of both (a) the kinds of things that could be done just by manipulating appropriate data sets, (b) that so much is possible if you just have the data.

Of course, having the data, or having a text in an actionable form, is itself half the struggle.

If you’re a totally beginner like me, and want to follow through those 7 tasks, go ahead, and feel free to drop me a line if you get stuck. There’s lots I don’t know, but I know enough to hint you along the path.

Project: Shepherding a text from print to digital

One of my projects for 2018 is to take a text and shepherd it, or curate it, all the way through an open source pipeline from ‘print’ to ‘digital edition’. This is part of my 2018 year of digital humanities. Here I talk a little bit about the envisioned process.

The text I have in mind is quite short, just over 2000 words. It’s Gregory of Nyssa’s De Deitate adversus Evagrium (in vulgo In suam Ordinationem). I’ve done some work on De Deitate Filii et Spiritus Sancti and this will be a nice complement to that.

My checklist of things to do:

The Pipeline

Step 1: OCRing a print text
Step 2: Correcting the OCR output
Step 3: Create a TEI-XML version.
Step 4: PoS Tagging/Lemma tagging/Morph tagging
Step 5: Produce a translation
Step 6: Alignment
Step 7: Annotations and commentary

Then, voilá, open-sourced text freely available with useful data attached. Half of these things I don’t actually know how to do yet. Maybe more than half. That’s part of the fun. And, presuming it goes well, will make it a pilot project for future texts through a similar pipeline.

Moving to project-based management: todoist and todo.vu

Since my life/work balance, and my work/work balance, has become more and more fractured, I’ve felt the need to organise it more effectively. Last year I started using todoist, which I love because there’s nothing I quite enjoy as much as ticking off a to-do list, and todoist has a ‘karma’ counter built in that racks up meaningless numbers as you tick things off.

This week I’ve started trialing todo.vu, and I’m now kind of using the two in tandem. It’s helped mentally organise the things I do, and todo.vu includes time-tracking in it, which is really helpful for me.

So, this post is just kind of a run-down of how I’m organising things and using these tools.

todo.vu has a hierarchy of Clients > Projects > Tasks.

This isn’t ideal, but I’ve just adapted it as seems good. todoist has Projects and Tasks, but Projects are endlessly nestable (well, I’ve never gone past about 4). So this is aligning nicely.

My Clients list looks like this (And my top-tier todoist now mirrors it):

  • Internal
  • Self-Education
  • (College)
  • (2nd College)
  • (3rd College)
  • Academic Career
  • Church
  • Online Tutoring
  • (A Business)


Internal was a pre-existing client, I’ve repurposed it for ‘life management’, whereas Self-Education is more specific educationally/professionally directed projects. The 3 colleges are institutions I do work for. Academic career is here I place things specifically academic: research papers, phd to monograph, etc.. Online Tutoring is self explanatory; if I were using this as a billing platform I’d create each person as a separate client, but since I don’t, I prefer this breakdown.

One of the things the whole schematisiation process helped me with was sorting through ‘recurring open-ended projects’ and ‘defined, goal-oriented projects’. For example, under Self-Ed I have separate project for “Latin studies”, “Greek studies”, “Gaelic studies”. These are all very open ended ‘projects’. Previously I had just set them as recurring tasks in todoist. Now, I’ve created individual Tasks for each component in todo.vu, e.g. reading Athenaze, working on a specific course, etc.. I can still set these as recurring tasks in todoist, (e.g. daily reading tasks), but it means I can log time on todo.vu, and when I finish an individual task, I can complete it while keeping the open-ended project going. So all book reading is also set up in the same fashion. Truly open-ended ongoing tasks, I’ll set on a year-by-year basis. For example, “Online Tutoring > Student 1 (2018)”, and just log hours against the task and close it at the end of the year. But, complementing this in todoist, I might also have “OT > Student 1 > this week’s prep” as an individual task.

I’m hoping that the net effect of this will be to give me greater awareness, control, and discipline over the disparate projects and tasks on my plate. That, combined with tools to limit/block-off various distractions, and here’s to a productive 2018. (I’ll just set a reminder now to blog on this again in 6 months…)



Diary of a Digital Apprentice (1)

One of my goals for 2018 is to acquire a working skillset in areas of Digital Humanities. As I do so, I plan to blog regularly on that ‘mission’. In today’s post, I provide some context for the start of that journey.


I’d say I’ve long had a user-side interest in Digital Humanities. I’ve appreciated, and used, the considerable resources that things like Perseus, TLG,  PHI, and other packages have presented. And I’ve always envisioned ‘more’ being possible. But, being relatively short on the technical side of things, DH has always been a bit of black-box wizardry to me.

A couple of years back I made the acquaintance, first digitally, of James Tauber. Some of our initial overlap and discussion had to do with tools for language learning and teaching. We met briefly at AARSBL in 2015, and conversed a bit more since then. Another face to face meeting at AARSBL in 2017 helped solidify things and we have launched both some collaboration, but also some apprenticing.

That ‘looks like’ two things. Firstly, a combination of push-learning, pull-learning, and hunt-learning. Pull, where I ask, “how do we do X?” or “is Y possible?” and then get a crash course on how to make certain things happen. Or an explanation of “yes, Y is possible, look, Dr ABC has been working on this for umpteen years, see!”. Push-learning is where you learn things you didn’t know you could learn, e.g. “Hey, Seumas, did you know  you can use E to accomplish F, G, and H!” And hunt-learning is when James says something like, “Seumas, figure out how to do M, N, O, and P, and then tell me how you did it or when you get stuck.”

Part of this relates to the work that Eldarion is doing on developing the Scaife Viewer for Perseus. Which is incredibly exciting because (a) Perseus! (b) have you seen the Scaife Viewer demo’d? (c) it’s great to see inside the black-box so to speak, to see how something like this gets developed and figure out how it works.

Another side of it is my digital Nyssa project for the year.

“Digital Nyssa” is my project to curate/shepherd a text ((initially just one, but maybe more)) through an open, free, digital pipeline from print to digital edition. It’s both a means of acquiring practical DH skills across a range of tools (OCR, TEI-XML marking, PoS and morph tagging, digital edition creation and then commentary/annotation/translation). You’ll be hearing more about it as the year goes on, and I’ll outline a little bit more next week.

So, each week I’ll be posting up a bit of what I’ve been doing/learning/working on, as part of a bigger project to self-document the learning process for myself, and hopefully encourage others that DH is not so scary. The first few weeks will play some catch-up too on things over the past few weeks.

Translation is not meaning

One of the downsides of training students to translate in order to understand, is that they very often develop the erroneous notion that translation is meaning. “The meaning of Greek word X is English word Y”, or slightly more complex versions of the same.

No, no, no.

Greek (or whatever language) means what it means, with reference to Greek, with reference to reality, with reference to its referents. Sure, I can concede that “Greek X means English Y” is sometimes just shorthand for “English Y is a suitable translation of Greek Y in this context”, but very often it’s not, it’s shorthand for “Greek X really means English Y, why didn’t they just write in English in the first place and make my life easier.”

Don’t fall for the trap. Figure out meaning first, then figure out how to render that meaning in your other language. That’s what translation is.

(I’m going to start trying to micro-blog more language/Greek/Latin/etc. mini-posts like this)

Can you teach a language with no native speakers via ‘immersion’?

A question I was asked recently on twitter.

Yes, but it depends what you mean by ‘immersion’ and what you think the goal is.

I don’t normally use the term ‘immersion’, because that normally, strictly, pertains to a method that involves full-time immersion in a language, so all day, all week. That’s simply not possible long-term, unless we set up little Latin-speaking (or Attic, or etc) villages. I suppose immersion is possible in short term venues like the SALVI Rusticationes

The goal, I would think, is also not to produce a new native speaking community. Though that is theoretically possible. See, the revival of Hebrew as a modern language, and also revival efforts of indigenous languages. But that’s not our goal for historical languages. If we revived a whole community of Latin speakers or Attic speakers, and allowed them to ‘continue on’ as a community and the language evolved, we would miss the point of learning ‘Latin’ or ‘Attic’, or ‘Koine’.

But there’s nothing inherent in Latin, or Koine, or Biblical Hebrew, that means they can’t be learnt as active, communicative languages. There’s nothing inherent in them that means people can’t become competent second-language speakers of these languages.

There’s a pedagogical challenge – how do you get people to this level, if few people are at that level? I think this requires careful research and considered pedagogy. We ought to take due effort to make sure that we are teaching language that is correct, as correct as possible, to what we know of the historical corpus. And we should recognise that teaching Ciceronian Latin, or Demosthenic Greek, or the like, is an artificiality. But it’s an artificiality that we already embrace in other methods anyway, and an artificiality that is desirable to the extent that we want people to read those kinds of texts and discuss those kinds of texts.

In terms of pedagogy, it’s about learning, and then using, those pieces of acquired language confidently and correctly. And perhaps the humility to say, “Well, I don’t know how to say X, Y, Z, let’s research together and get it right.”

Grammar/Translation, Communicative methods, and comparisons…

This is me picking up some pieces and adding nuance to one of my usual hobby horses, that is communicative approaches to ancient/classical/historical languages.

  1. Does Grammar/Translation ‘work’?

I think this partly depends upon what we mean by ‘work’. My main interest is developing students who can read effectively in their second language – without translating in order to understand, and mentally operating in the second language.

I don’t think G/T normally produces this for most students, or even most students who respond well to G/T. It seems to me that G/T produces primarily students who translate in order to understand (as opposed to translating messages they have already understood). I don’t deny, that some long-time practitioners of G/T end up being very competent readers. I’ll get back to that below.

  1. Are communicative methods better than grammar/translation?

It’s hard to make a proper comparison, because really this requires a controlled, data-driven study in Second Language Acquisition, and while there is obviously some work going on, on that question, I have yet to read a full-blown comparative study on the question.

However, everything I’ve seen in SLA suggests that G/T isn’t ‘the way to go’. And given that G/T dates back about 200 years, and has been largely abandoned in modern language programs, and the contrast between G/T products and modern language programs, I think we need to consider that pound for pound, G/T is not the best method.

  1. Is it far to compare seminary programs with classics programs?

No, not really. But that’s because a typical seminary program fits Greek grammar into a single year, and then (if you’re fortunate enough to have a robust program) 2-3 years of New Testament texts in Greek. This is ¼ to 1/6 of a program. And it’s focused on intensive, not extensive, reading; on analysing texts at a micro-level and exegeting for meaning; on accessing technical discussions at a verse-by-verse level.

A classics program is 3-4 years of language and literature in that language. If it’s a traditional classics program, that’s at best a 50/50 split (if you do nothing but lang and lit in Greek/Latin). Having done a full Latin sequence, I have a fair idea of what that looks like and produces. Yes, classics produces better readers, but largely because it’s a lot more exposure to texts in the target language.

And this, I’d contest, is ultimately why G/T produces readers – not a superiority of method per se. Translating is a way of making a target language message comprehensible, and sheer volume of comprehensible input is what produces language acquisition over time. G/T will do that. I’m just not convinced it does that as effectively as it could.

My main push-back to G/T is driven by the fact that G/T has this “grip” on both classics educators, and biblical language instructors. There’s a conservatism that thinks G/T is (a) the way it’s always been, and (b) the tried and true method, that (c) ancient languages are ‘different’ and cannot be taught communicatively, (d) etc., etc..

I have plenty of appreciation for grammar, for translation, for G/T methods, but I have no appreciation for traditionalist views that are ignorantly dismissive of alternatives, and critiques, of G/T.