OLDaily

English isn't generic for language, despite what NLP papers might lead you to believe
Emily M. Bender, Symposium on Data Science & Statistics, 2019/12/26

This is from last March, but I found it today, and the title alone is worth passing along this set of slides. "Natural language isn’t just English, and NLP work should stop pretending that it is. If you’re a consumer of NLP tech (e.g. for text as data research), demand better." See also: Wenyan, "an esoteric programming language that closely follows the grammar and tone of classical Chinese literature. Moreover, the alphabet of wenyan contains only traditional Chinese characters and 「」 quotes, so it is guaranteed to be readable by ancient Chinese people."

Web: [Direct Link] [This Post]

Undergraduate Philosophy Journal of Australasia
2019/12/26

This link is to the first issue of the Undergraduate Philosophy Journal of Australasia. I like not only that it is open access, but also that it is by and for undergraduate students. Now I have my criticisms of journals and that whole practice, but I also think it's really important for students to engage in the actual practices of a profession (for better or worse) as they learn about it. Also, it gives us an insight on what they study and find relevant - for example, about 'vice-charging' as "a process whereby individuals or groups can collaborate to address epistemically vicious behaviour."

Web: [Direct Link] [This Post]

Educational visions
Rebecca Ferguson, Ann Jones, Eileen Scanlon, Ubiquity Press, 2019/12/26

I spent the better part of Boxing Day afternoon reading and mostly enjoying this book (186 page PDF). It is based on the work over the last 40 years of the Computers and Learning Research Group  (CALRG) at the U.K.'s Open University. The point of departure is CALRG's "Beyond Prototypes" which is used to explain "why educational technology initiatives worldwide succeed and why they often fail." This then informs  four major areas of inquiry: teaching and learning at scale, accessible inclusive learning, evidence-based learning, and STEM learning. Each is given a historical perspective, then in a separate chapter a look forward. In a commentary on the book Martin Weller describes it as a "good example" of an alternative to the "wilful historical amnesia in much of ed tech." Maybe so. But let's not forget that this is a book specifically about the Open University, and that while nobody doubts the OU's importance to the field, nobody would say that it alone defines its history, despite the often subtle ways the book says just that. Still. It's a good read, well worth the time.

Web: [Direct Link] [This Post]

The Five Stage Model
Gilly Salmon, 2019/12/26

Gilly Salmon is asking for feedback on her five stage model of online learning. I'm not a fan of 'stages' - the idea that we would been socialization and information exchange before knowledge construction and development seems overly formal; all of these happen all of the time at once. And these days, we don't need to 'access' so much any more; we just turn on the phone. Also, we rarely socially 'construct' knowledge; it's a much more organic process. Some interesting comments on the LinkedIn post; for examply, Jeremy Main says "normalisation, social integration, creates a society devoid of ambition."

Web: [Direct Link] [This Post]

How do you teach a car that a snowman won’t walk across the road?
Melanie Mitchell, Aeon, 2019/12/26

I understand the concern, but I think it relies on a myth that won't - in the long run - bear scrutiny. The argument here is essentially that human drivers depend on a wealth of knowledge known as 'common sense' - "the mostly tacit ‘core knowledge’ that humans share – knowledge we are born with or learn by living in the world." The myth here is that there is 'core knowledge', that it is common, and that it is required by an autonomous car. The sort of person who uses 'common sense' to plough through a pile of leaves or into a flock of pigeons is the sort of driver who has accidents because "nobody could have predicted" the hidden tree trunk or the damage a pigeon can cause. The AI, meanwhile, will learn a very different 'common sense', one focused on road safety, and hence not be misled. Via The Next Web.

Web: [Direct Link] [This Post]

The Answer for Schools Is Not More Technology. It’s Teachers and Human Connection.
Danielle Arnold-Schwartz, EdSurge, 2019/12/26

This item hits all the right buttons but it's fundamentally wrong. Yes, business-minded people thing of ed tech as a way to save money. Yes it's true that in many cases people need a human connection. Yes it's true that education is a core requirement for democracy. But it does not follow that "We need educational technology that puts highly trained teachers at the center of product design and implementation." And it does not follow that we should "forget the efforts to appeal to fiscal reforms." We need to lower costs, if education is to reach seven billion people. And there are many ways to create the necessary human connections without requiring teachers - our friends, family, co-workers, peers, mentors and online community are all made of people, and in many cases in a much better place to help us learn. Teachers can be important, but they need to go through their own Copernican revolution.

Web: [Direct Link] [This Post]