#computational linguistic

[prokopetz]

Tried an LLM-powered thesaurus today just for giggles and it helpfully informed me that a synonym for "leg" is "penis".

[official-linguistics-post]

A new look at our linguistic roots

Linguists and archaeologists have argued for decades about where, and when, the first Indo-European languages were spoken, and what kind of lives those first speakers led. A controversial new analytic technique offers a fresh answer.

An article from early 2024 on the use of computational phylogenetics for historical linguistics work on Proto-Indo-European. Read at Knowable Magazine (no paywall).

[elbiotipo]

I'm not a linguist and I find the whole excercise of conlanging, while I love it and respect it, beyond my abilities, but I do have one thing or two to say about linguistic diversity and how boring is to have a "common" or "basic" language in fantasy or science fiction without exploring the implications.

Being a bilingual speaker of Spanish and English, and someone that because of work reasons and entertaiment tastes interacts a lot with English, I tend to see English as the equivalent of those "common" or "basic" languages of speculative fantasy. As a useful tool for communication, science technology and commerce. In real life, however, as you are aware, the expansion of English tends to undermine local languages, it's considered more valuable to know English that to know the language of your grandparents, or learn any other language you just feel curious about.

The experiences of every multilingual person are different, but in mine I know English, I write and read and listen to English a lot. But I don't consider myself an English *speaker*, I speak Spanish and more to the point Argentine Spanish, that's the culture I identify with, and it's the language I use to express my feelings and inner thoughts. I can't imagine saying "I love you" to anyone in English, to me it's just a tool I use to access to knowledge or communicate through language barriers ("basic", "common"). But interestingly, by both writing and participating in the wider English-speaker internet culture, isn't it part of my own culture, as an individual, too?

The fact is that English also has a culture(s) and a history and a corpus of literature. So when we write about "Common" or "Basic" languages in fiction we need to ask ourselves: where did they come from? How did they become the standard? Is there a literature, a canon, a culture of "Common" in your fantasy world? What about other languages, other cultures that aren't raised learning it and see it just as a tool? Because no matter the strenght of Anglophone cultural imperialism and the social value of learning English, I don't see Argentines, or for that matter Chinese, Italians or Russians abandoning their first language. And yet even in English and in all other languages (ESPECIALLY other languages, English is remarkably uniform) there is a variety of dialects. And we need to remember, once Latin was spoken only in a village in central Italy, and English in a rather remote rainy island. They weren't destined to have their future roles, history drives language.

So, when an author goes for the "universal language" explanation to avoid linguistic misunderstandings, for me, it raises more questions that I believe are worth exploring.

[savagegardensprogramming]

The Purpose of Language in General

I found that Language has three important uses is our every day life.

(1) The first one is obvious. Language as means of communication. We use language to communicate our ideas and desires to other people. duh!

(2) The second one is more subtle but also just as important. Language as means of articulating thought. We use language to construct and hold ideas together. Even if we do not like talking to people, language is incredibly important.

Ferrel children are children who grew in the wild with no access to human upbringing. These children are mentally handicapped and I believe this is largely due to lack of language. 10,000 years ago our brains where primitive but not very different from today. My personal hunch is that crude language skills leads to crude thinking. Thats how we got by in those days. And that is the difference between us modern humans and humans 10,000 years ago.

There are people who do not have an inner monologue yet they are very intelligent. I would bet that these people still have some sort of mechanism similar to language that allows them to articulate thought. Its based on other brain machinery such as vision and emotion.

(3) The third one was pointed out by a colleague. Language is a window to outside observation. We don't have to be involved in the conversation to understand what other people are talking about and learn from them. Documentation important also. History is preserved by written language. Only after written language did humanity started to thrive.

Sooo, what does language have to do with computer science, you may wonder. Well, programming languages are real languages. My claim is that programming is a communications problem. I see the machine as an alien species from ourselves. And the task at hand is to find a way to communicate to them and inform them what we want them to do.

(1) The first aspect (the obvious one) of programming languages is to do just that. We need to communicate to the machine what we want them to do.

(2) The second aspect (the more subtle one but just as important) of programming languages is for our sake to understand what we are telling the machine. We use programming languages to articulate ideas in our head before typing it on the computer. That is why we insist and persist on using JSON and XML to store data. Even thou those formats are incredibly inefficient from the view point of a machine.

(3) The third aspect of programming languages is a form of documentation. Other people can understand what the programmer is telling the machine. Its why open source is popular among programmers.

[his-dark-materials-trilogy]

Etymology of daemon from Oxfrod Languages

[front-facing-pokemon]

[noosphe-re]

virtual (adj.)

late 14c., "influencing by physical virtues or capabilities, effective with respect to inherent natural qualities," from Medieval Latin virtualis, from Latin virtus "excellence, potency, efficacy," literally "manliness, manhood" (see virtue). It preserves Middle English virtue in the sense of "efficacy, power to do."

The meaning "being something in essence or effect not in fact; existing virtually though not actually" is attested from mid-15c., probably via the now-obsolete sense of "capable of producing a certain effect" (early 15c.). Opposed to actual, real, literal.

By 1831 in optics, of apparent images, as in a reflection. The computer sense of "not physically existing but made to appear by software" is attested from 1959, originally of memory.

Etymonline

[weird-machine]

That's a nice solution you've got there. It would be a shame if it wasn't the root cause of the problem.

[dramatic-dolphin]

there's two kinds of university professors and they are "i take attendance every class, if you're sick you need to tell me a day before, you can only miss three classes per semester" and "hey i noticed you haven't been to class at all in the past two months, are you handling the coursework okay? feel free to ask me if you have any questions!"

[withercat22]

A friend and me had a conversation, which led to us wanting to confirm what we were thinking. Our "hypothesis" (very informal one xd) is that there's a correlation between people who study languages, and those who are learning programming. As such, here's the poll:

Please, reblog for a larger sample size!

[pizzaronipasta]

Computer Cursive

Okay so, I've had an idea. Y'know how the reason why the world's writing systems look so different is often because they were developed for writing on different materials? Well, to my knowledge, there hasn't ever been a writing system developed from the ground up for handwriting on a computer using a mouse. Which is a shame, because other writing systems are so ill-suited to mouse handwriting.

The Latin alphabet was originally carved into stone. But then a cursive script came around so it could be written on paper more easily. Who's to say the same can't happen with mouse handwriting?

So I tried my hand at making my own "computer cursive." Since precision with a mouse is much harder than with a pen, chisel, or stylus, I tried removing as much fine detail from the letters as I could while keeping them distinct and recognizable. Specifically, producing strokes of precise shapes is especially hard, so I prioritized making the arrangement of strokes recognizable, no matter how the strokes themselves ended up looking. Also, since long continuous lines are less than conducive to digital error correction (ctrl+z would remove a whole word instead of just a single letter or less), and because the risk of making an error gets quite high after only a short while of continuous marking, the term "computer cursive" ended up being something of a misnomer. It was best to keep letters disconnected from each other, thus not fulfilling the definition of a cursive script. However, "computer cursive" makes for a convenient shorthand that's reflective of the thought process behind its creation, so I'm sticking with it.

Anyway, here's what I came up with:

So first of all, it looks cool as shit if I do say so myself, though that's probably just because of the calligraphy pen brush I used (and I recommend you do the same if you ever try using this). But in seriousness, I'm very happy with how easy it is to recognize most of the letters here, and the rest are easy to get used to.

Some usage notes:

the letter c should start going diagonally down and left, and gradually curve further downwards. it doesn't matter how far it ends up curving, as long as the curve is noticeable

the letter k needs its bottom-right stroke to be curved as seen in the image, or it'll look too much like an h

yes, the letter o is a triangle. deal with it

p and q are distinguished by the lengths of their rightmost strokes: p's is longer (roughly as tall as the other stroke) and q's is shorter

r and v are distinguished in the same way: r's is longer (taller than the other stroke by a noticeable margin) and v's is shorter (more symmetrical). also, r's stem should technically be vertical while v should technically be completely symmetrical, but that isn't a feasible distinction to reliably make in practice, so don't worry about it

the letter s should start going diagonally up and left, and gradually curve down and around to start going diagonally down and right. it doesn't matter how far it ends up curving, as long as it fits this description

the letter u should start going straight down, or down and slightly right, and gradually curve further right. it doesn't matter how far it ends up curving, as long as the curve is noticeable

In general, don't worry too much about it looking tidy—it's specifically meant to be serviceable despite inevitably looking sloppy from time to time.

[official-linguistics-post]

New open-access article from Georgia Zellou and Nicole Holliday: "Linguistic analysis of human-computer interaction" in Frontiers in Computer Science (Human-Media Interaction).

This article reviews recent literature investigating speech variation in production and comprehension during spoken language communication between humans and devices. Human speech patterns toward voice-AI presents a test to our scientific understanding about speech communication and language use. First, work exploring how human-AI interactions are similar to, or different from, human-human interactions in the realm of speech variation is reviewed. In particular, we focus on studies examining how users adapt their speech when resolving linguistic misunderstandings by computers and when accommodating their speech toward devices. Next, we consider work that investigates how top-down factors in the interaction can influence users’ linguistic interpretations of speech produced by technological agents and how the ways in which speech is generated (via text-to-speech synthesis, TTS) and recognized (using automatic speech recognition technology, ASR) has an effect on communication. Throughout this review, we aim to bridge both HCI frameworks and theoretical linguistic models accounting for variation in human speech. We also highlight findings in this growing area that can provide insight to the cognitive and social representations underlying linguistic communication more broadly. Additionally, we touch on the implications of this line of work for addressing major societal issues in speech technology.

[drlinguo]

Language and computers (Open Access)

Lelia Glass, Markus Dickinson, Chris Brew, Detmar Meurers

This book offers an accessible introduction to the ways that language is processed and produced by computers, a field that has recently exploded in interest. The book covers writing systems, tools to help people write, computer-assisted language learning, the multidisciplinary study of text as data, text classification, information retrieval, machine translation, and dialog. Throughout, we emphasize insights from linguistics along with the ethical and social consequences of emerging technology. This book welcomes students from diverse intellectual backgrounds to learn new technical tools and to appreciate rich language data, thus widening the bridge between linguistics and computer science.

[abyssalsquiid]

linguistics has grown so much as a field in the past 20 or so years, you'd think we could do better than fucking treeform

[yuritypology]

'Optimality Theory is noncomputable by Turing Machines'

Another Penrose win

[aphel1on]

really all i should need to do to get professionally dx'd with autism is show them my warrior cats conlang from senior year of high school