Category: Random

When In Rome, Do As the Linguists Do

Linguists will tell you they don’t study the correct way to speak a language, but rather the way it is (empirically) spoken. The rest of us, however, do abide by the rules (grammar, semantics) and view them as a social norm to be observed.

It is hard to pinpoint the exact reason why speaking a language ‘correctly’ is necessary or desirable; though one could single out social constructs as a reason to disfavor those who fail to ‘speak correctly’, hard objective truths as to maintaining correct form are hard to come by or prove. After all, we all know any modern language is basically a corruption / variation of older languages. Last decade’s mistakes are today’s rules, so is there a point to making a fuss about people’s mistakes?

Two easy examples are הַנְגָּשָׁה and literally.

The first, הַנְגָּשָׁה, has always annoyed me personally. (If you don’t speak Hebrew, I would skip this paragraph). Hebrew verbs are constructed via root and conjugation. The root נ.ג.ש. in the binyan of הפעלה already exists, namely הַגָּשָׁה. The nun was ‘swallowed’ over time into the gimel (explaining this is in English is kind of ridiculous); this is the meaning of גזרת חסרי פ”נ. In the fact, the meaning of הַגָּשָׁה is literally (snicker, see next paragraph) הַנְגָּשָׁה – to make something more accessible. For some reason in this present, post-modernistic world, people feel they have to make up a new word to convey a pseudo-new meaning. Well, the meaning isn’t new. “Making accessible” does have a modern liberal meaning relating to helping disabled people – be they blind of sight, have a language barrier, or in need of special physical access – but the core meaning (as in, “make this accessible to someone”) has always existed, even in the simplistic version of could you pass the salt? == make it accessible to me?, which is of course a common use of the word הַגָּשָׁה. So we don’t need this new, contrived creation of הַנְגָּשָׁה: we already a/the word for that (and it’s basically the same word).

The second example, one which might be more familiar to the English-speaking world at large, is the use of literally – or rather it’s misuse, as in “the movie was so good, I was literally glued to my seat.” To many people’s chagrin, this word seems to be gradually losing its meaning. I was enlightened to recently read on Quora a humbling perspective on this, in the same vein of the above – many words gradually lose their meaning, so no need to get worked up about just another word. The prime examples would be reallywhich literally means/reads “in reality” (and thus should be used the same way ‘literally’ is, but in effect is of course often used just to emphasize) and very (from the Latin root for ‘truth’; should be used for in truth but is of course just another tool for emphasis). So it seems literally may just be going down the same path – or more importantly, we use so many words in a non-literal sense that it doesn’t make much sense to get worked up about one more.

Despite any such objective legitimacy, however, it is hard to respect someone with a faulty grasp of their language. It seems the standard is social, rather than absolute; ‘correct’ language would such as spoken and understood by those around you – including any arbitrary rules observed or neglected.

Perhaps as in when in Rome, then – when speaking a language, do as the linguists do.

Baruch’s Choice (Of Where To Go To Grad School)

My friend Baruch is debating where to go to graduate school.

Baruch is close to finishing his BSc, a triple-major in CompSci, Math and Psychology at the Hebrew Univ. His grades are good – around an average of 90 (in each and in total), and grades nonwithstanding, he is one of those rare people who ‘get’ CS, feel both comfortable around it and genuinely interested in it. He reads Math StackExchange for fun. He is border-line Dean’s List student. On a personal note, Baruch and I were the only two students of our grading majoring in Psychology-Computer Science together (he added Math mid-degree).
Unfortunately, Baruch is still in the undergrad mindset (fostered and encouraged in no small part by the science faculty) that he is insignificant to the all-awesome Institution. This is a mindset which (I had shared, and) I assume should change quite quickly after completing his BSc – but it will reflect on the decision regarding where to go to grad school for, a decision to be taken soon.
Baruch has already decided he is going to continue with CS studies (instead of going to grad school for something else, or getting a day job, or becoming a Nepalese sherpa) which significantly narrows the conundrum. I think this is the right decision for him.
By Baruch’s own admissions, he has narrowed it down to three main options (and I love writing this, because what seems like a simple description shall probably help influence the discussion itself; see framing):
  1. HUJI – Pros: 1. Jerusalem is a friendly, “known” environment, both in terms of the University and socially (though this might change in a year or two, as friends finish their own undergrad and move on). 2. Possibility of taking graduate classes during last semester of undergrad. Cons: 1. Spending too many years in the same city. 2. Rejuvenate and diversify both social circles and academic environments. 
  2. Technion – Pros: 1. Haifa is a real friendly, liberal (and cheap) city, 2. diversity of academic circles. Cons: New social circle will involve effort; academic environment is unknown. 
  3. Weitzmann Institute – Pros: Considered best in Theoretical Computer Science (Baruch’s field of choice). Cons: unattractive social scene.

It is hard to encapsulate in ‘bullet-point form’ the various associations with each choice. Any of the above, after all, will also determine which city Baruch will live in for the next couple of years. As a full disclosure, I would naturally be happy if Baruch chooses to live in Haifa (where my parents live and where I currently live) as it would make it easiest for me to see him.

My personal opinion is that the best option would be the Technion. It’s a good academic institute – but they all are. In my view, the main consideration to take into account should be everything other than the academic institute itself; that is, studies aside, where would you want to live for the next couple of years? Having said that, my preference is towards diversity over any known comfort: life is short, go and experience living in many places. That should rule out HUJI. I know Haifa, and I know it is a great city for someone like Baruch – not too big and flashy, very liberal, big student population, etc. 

Difficult decisions. Perhaps Baruch should do as I did, and let the winds of fate decide for him.

Greg – Back-of-the-Napkin Calculations

The “Greg” coffee chain was founded in Haifa (in 1994, according to their website). Some ten years ago, they had three branches, spread out across about  a mile of the Carmel. One of the branches was the closest coffee-shop to my parent’s house, was open 24/7, and had a cozy, laid-back atmosphere – so I found myself frequenting it often.

One of those nights, I was sitting with my friend Orr (who spells his name with a single r, as in the conjunction, but I maintain it should be spelled with two r’s, both to distinguish from the preposition and as in the character from Catch-22 (Wikipedia entry)). We idly got around to calculating how much money Greg were making. It seemed to me – based on the back-of-the-napkin calculations to be presented below – that Greg should making a fortune.

Now, Greg are open 24/7. Save for 4 days a year (Passover, Memorial Day, Holocaust Day, and Tish’ah be-Av), let’s assume month-round (30 days a month). The Kikar-Sefer branch we are talking about hosts some ~50ish tables for two (some lumped together to form a 4-6 sitter, of course). An average table of two should pick up an hourly tab of no less than 50 NIS (probably way more, but let’s stick with conservative estimates). Greg is never empty; even around 4 a.m. you would see some 4-8 tables occupied, standard hours would show about a 25% occupancy, and peak hours are completely booked, of course. (Greg being a popular venue on prime real-estate) My estimation — and this might be a key arguable metric, of crouse — would be an average occupancy of no less than 25% – i.e., some 12 tables occupied per hour on average, for any given hour, all day long (all week long, all month long).

Using the above assumptions, this would result in an average hourly revenue of 50 NIS per table, times 12 tables, == 600 Shekels an hour, or 14,400 Shekels a day, or ~430,000 a month. Assuming a staff of ~12 workers, each at an assumed  input of ~5,000 a month, would take a bite out of that at 60,000 – bringing profits down to 370k. In a 50-NIS tab presumably the order consisted of nothing more than drinks, whose ingredients’ prices are negligible. (If the tab includes something more pricey such as vegetables or meat, the tab would be significantly higher – in our conservative estimate we are ignoring those cases, but they only boost earnings higher.)

Situated on prime real-estate, let’s assume the rental costs are a whopping 100k per month – bringing profits down to some 270k. Assume another 100k in utilities and other arbitrary expenses – we’re down to still 170k of profits per month. This figure was reached what seemed to me to be quite conservative estimates.

Each assumption could be argued separately, of course, but I fail to observe any estimate (or combination of estimates) that would throw the end result off in an order of magnitude, even if changed drastically. Perhaps the most contentious claim is the 25% average occupancy rate, though having personally frequented the place at pretty much every hour in the day, I can say from first-hand experience it shouldn’t be much less.

So 170,000 Shekels in profit per month, on average. Which would work out to about 2 million shekels in profit per year(!). Big numbers.

Back to the story, when we finished with these back-of-the-napkin figures and reached what we thought to be an astounding result, we asked to see the manager/owner. He was a young guy – probably not much more than 30-something, and amiable enough. He wouldn’t confirm the end result, but acknowledged most of the estimates were close enough.

Today, Greg has 84 branches across Israel. Maybe the napkin’s numbers were close enough.

(Further reading, not immediately relating to post’s content: Quora/Forbes on how to run a successful coffee-shop.)

Coursera: World-Class University Courses Online

My recent favorite site is Coursera.

In a nutshell, they team up with leading universities to create online courses, each comprised mainly of short video snippets of the lectures. You can ‘enroll’ in a course (one of various subjects) and take part in ‘quizzes’ as part of a pre-defined timetable, or you can just watch the videos whenever you feel like it.

Mock timetables are as expected quite efficient and popular in helping people commit to study plans; previous efforts to provide online courses (such as Stanford’s Databases course and Machine Learning course) offered not only a set timetable and a Q&A forum and exercises, but even an auto-checking and -grading mechanism for the exercises, to further incentivize the participant and allow feedback loops. Auto-grading is almost expected when dealing with a technical matter (where the exercises often result in some actual code being written, which can easily be tested) but I would argue than any subject can be formalized into an auto-testable format, at the very least in the form of a multiple-choice test. AFAIK the courses offer either mock credentials upon completion or none at all, so there is little reason to cheat. (Although plagiarism does exist sometimes; see e.g. here)

I’ve always been interested in teaching techniques, and it’s fun to see this huge advancement. This is a step forward from just “let’s put a webcam in the classroom during the lecture and then post the video online”; the snippets are each 5-15 minutes long (short enough to easily concentrate on), and most of the time the video is centered on a dedicated whiteboard displaying slides on which the lecturer can draw (with a special pen), displaying his comments and marks immediately. Pretty much the way I’d want it to be like… (a transcript of the lectures would be good, though, as many people read faster than they listen, certainly when ingesting a new subject which requires reading multiple times).

Coursera certainly aren’t the first in their field — Stanford (and others) had already been offering some form of online courses, Khan Academy is better-known, and Udacity is similar. Slowly but gradually things are changing, and one day even universities in Israel might have to change their arcane ways of teaching everything orally “live”.

On a personal note, I’ve been trying to get to study Andrew Ng’s Machine Learning course on Coursera. Updates on that as events warrant.

So anyway, go on and check it yourself — no excuses now for a bad education other than your own laziness. 🙂


Moving In

Limor and I moved in together the other day; namely to my parents’ housing unit alongside their house. Limor and I have already effectively spent the last couple years sharing living quarters, but that included sharing them with various gracious roommates, and this is the first time we’ll be on our own. Using my parent’s housing unit gives us a comfortable independence from the fiscal burden, and more importantly, from the long-term commitment. In our particular case, we are not sure yet where we want to live for the next year or so, as this also depends on the results of our current job-hunting. So for the meantime, we can at least start enjoying living together, although our wandering (better conveyed as “nedudim”) will likely continue in the months that follow.

No Rest For the Wicked

No rest for the wicked” (or “No peace for the wicked“) is a phrase originating from the Book of Isaiah verses 48:22 and 57:20-21. (No rest for the wicked – Wikipedia, the free encyclopedia)

The Biblical translation to “rest” is interesting, since the original Hebrew word (as Wikipedia points out) is “Shalom”, which is more commonly used as “peace” or perhaps “well-being.” I suppose the concept of the Lord (and us, his faithful minions?) constantly pursuing ‘the wicked’ is romantic enough to grant picturesque gravitas befitting the Anglo-Saxon interpretation. (“Smite” is another word that succeeds in the same manner.) Reminds me of Batman Begins’ “Death does not wait for you to be ready!”, which might as well have come from the Bible.