Supersonic Man

September 15, 2022

search engines used to be dumb, but now they’re stupid

Filed under: computing — Supersonic Man @ 8:16 am

What is the difference between dumb and stupid? I’m using the word “dumb” here in a specific sense, as a jargon term which is common among engineers and coders. A “dumb” device is one which does one simple thing and does not try to do more, while a “smart” one is encumbered with automated features and add-ons. A smart phone, for instance, tries to do as many things as possible, whereas a dumb phone only makes telephone calls. A smart microwave is able to automate tasks like defrosting meat and popping popcorn; to enable this it has 27 buttons, half of which are useless without repeated study of the manual. A dumb microwave, on the other hand, can be operated by anyone the first time they see it, because all they need to do is turn one knob.

Sometimes smart is good, and other times dumb is actually the better option. I want my phone to be smart but my microwave to be dumb. Hand tools such as crowbars and can openers are dumb, and that is a virtue. Smart devices are usually faddish, temperamental, and rapidly obsoleted; dumb ones may still be valuable or coveted a hundred years later.

Dumbness is relative. Back in the days of timeshare computing, for instance, it was common to access systems remotely with a device that combined a video display and a keyboard, which was called a terminal. As microcomputers started to appear these quickly came to be called “dumb terminals” because they did not contain their own computers. And yet these devices were quite complex, and would be considered smart if compared to, say, a typical television set of the time. Similarly, the dumbest telephones commonly in use today — the wireless handsets favored by those who still use landlines — would be considered smart in comparison to a traditional rotary-dial phone.

So even the earliest and most primitive search engines were hardly dumb in an absolute sense. They were, for their day, marvels of software engineering and database architecture. But they were still “dumb” in the most important sense, because they did one specific thing, and that thing was clearly understood and predictable even for fairly nontechnical users. They found documents according to their usage of the words or phrases you listed as search terms.

Of course those early search engines left plenty to be desired, so people quickly started trying to make them smarter. They came up with algorithms to better assess the relevancy of documents, and how authoritative they might be. This was based on techniques such as considering synonyms of search terms, and counting how many other documents made some kind of reference to the one being evaluated. And these helped. Searches got smarter.

But them someone decided that the time had come to switch from trying to make search smart, to trying to make it intelligent. That is, they tried to make it interpret the meaning and context of the words in users‘ queries, in the way that someone who understood the language might, instead of just taking them as literal. And they tried to make them respond usefully to questions people ask when speaking in natural language to each other, which requires treating some words as relevant search terms and others as just grammatical glue.

The first to make the attempt was Ask Jeeves. They said they‘d have a search engine which could understand and answer queries in English the way you would. They failed miserably at this goal. (Despite this, the company survives today as, and is even making something of a comeback.) The first to have some success was Google, and they have been the dominant search engine ever since, even though many others now offer a similarly capable service. They started adopting the techniques of artificial intelligence research, such as neural networks and machine learning. These techniques create a system that in a sense acts more through a sort of intuition than through algorithms — a system that can form judgments without either it or its creators knowing why it feels the judgment to be correct. They can only say that it seems like a good bet based on past experience in similar situations. Getting good results from systems is in some ways less like building an engine than like training an animal.

And it works brilliantly… up to a certain point. Most users do often ask it questions in semi-natural language nowadays, instead of thinking in terms of just locating relevant words. And the results they get, for common questions and general knowledge, are amazingly spot-on, though they may be polluted by ads, spam, and misinformation. But when a user, encouraged by this success, tries to take it further, and push a bit against the limits of that domain of success on basic topics, it’s pretty easy to reach a point where it starts falling apart.

Broad general knowledge and popular information work well, because you can learn rapidly what the majority consider a good valid answer. But when you start to dig into finer details, or look for critical analysis instead of consensus, or enter areas of specialized knowledge, that kind of automated learning no longer gets the job done. That’s when you run into the deep difference between artificial intelligence techniques and actual intelligence. That’s when search engines turn stupid.

If they were dumb, you could find ways to get good results out of them — you could try different vocabulary, emphasize different terms, use operator terms like AND/OR/NOT or NEAR/AROUND or an asterisk at the end of a partial word, and eventually work toward good results by a semi-predictable process. But because the engines are stupid instead of dumb, they don’t let you explore these options, but instead try to interpret these variations as “intelligently” as they can, which often means they just keep returning the same results for the different variations, undoing your efforts to differentiate your new search from the failed one. They try to help and save effort, but the result is to actively obstruct and thwart anyone who is trying to get past their wrong answers and bad assumptions — almost like the experience of dealing with a live person who is stupid, or aggressively ignorant.

And there’s no escape anymore. I try different search engines and they all act the same. I also try the old techniques that remain from the dumb days, like quote marks around mandatory words, and in a lot of cases they no longer work, even though on paper they’re supposed to. Google claims to this day that quoted terms still force an exact match, but I call bullshit. I see them break that promise constantly. It usually helps with phrases, but if I want to make sure of a single word, it often does more good to repeat the word three times than to put it in quotes. Using a dash to negate a word also doesn’t work with any consistency.

I would be fine with all the half-assed AI if there were a way to override it dependably, but it just takes the overrides as suggestions subject to interpretation, and fills the result page with irrelevant spew. What I want to see is some mode of operation in which I can enter four words, and be dead certain that I will see no results except those that really do contain all four. It’s surprising how often I need this. It seems like every week or two I’m doing some search in which leaving out just one word makes the whole page irrelevant. And I have yet to find any modern search engine that still supplies such a service.

As I’ve pointed out before, this loose behavior introduces biases into search results. The strongest bias is that it favors the broadly general over the narrow and specific. Attempts to steer it toward finer details get countersteered right back to generalities, unless you find specialized jargon terms that allow the detail in question to be differentiated by a distinct vocabulary. There’s also a bias toward the popular over the niche, and the consensus over the contrarian. Combine that with the biases introduced by advertisers, and it can be remarkably tricky for a search that includes the word “review” to land on an actual review of a commercial product or service, as opposed to a sales pitch or a spam page. Add those biases up, and the general tenor of search results leans toward passive conformity… except for the other bias, which is toward eye-catching clickbait and provocation, which helps magnify the ideas of kooks, paranoids, and bigots. Without apparently intending to, what Google and its imitators seem to be encouraging is a world in which everyone is either a happy consumer or an angry troll… or a greedy scammer who has found one of the many weak points that allow them to misuse search engines to exploit people.

What gets left out is nuanced thought, careful research, unbiased investigation, and forward-looking ideas — the things we most need to include in the discourse. In short, this attempt at artificial intelligence has ended up becoming hostile to actual intelligence. And you’d think that the AI tech would rapidly improve just like any other tech, but you can’t cure stupid that way. Artificial intelligence, as a field, is notorious for coming up with advancements that are very impressive up to some limited point, but then hopeless at trying to take a next step beyond that point, no matter how much more work is applied to improvement. If anything, the pace of advancement lately has only been digging us deeper into this hole. It may not get out of that hole until the search engine servers start achieving actual intelligence… at which point we’re likely to have much bigger problems than bad search results.

Since that very unpredictable breakthrough and its unguessable consequences are still many many years away, the result is that we’re now within a lengthy period where the services we rely on are stuck in an uncanny valley between being smart and being intelligent, where they can’t do either correctly. And there doesn’t appear to be a lot of hope for better alternatives to succeed commercially.

June 20, 2022

the meanings of “meaning”

Filed under: Hobbyism and Nerdry,thoughtful handwaving — Supersonic Man @ 8:01 am

I recently watched a video by Kyle Kallgren where he talks about Umberto Eco, which means talking about semiotics, which is the philosophical study of signs, symbols, and meaning. And while I definitely enjoy and appreciate Kallgren’s work, there’s a side to the meaning of meaning that he didn’t really mention or go into. I’m not sure if the big name semioticists he talked about really did either. And the aspect of meaning that I think got missed has to do with the kind of meaning that Douglas Hofstadter talked about back in the seventies in Gödel, Escher, Bach: meaning that inherently emerges because the signifier has an isomorphic relationship to that which is signified, like the way a road map relates to actual roads. The map conveys true messages about how the roads connect, and that meaning can be decoded even by someone who’s never seen a map before, once they observe enough of the roads to notice the matching patterns.


April 23, 2022

Bitcoin is a Ponzi scheme

Cryptocurrencies were supposed to be a new tool of commerce for people who don’t trust governments, not an investment commodity or a get-rich-quick scheme. Yet that’s what they’ve turned into. And the further we get into it, the harder it is to see the whole idea as anything but a con.


February 27, 2022

The Investigation of Consciousness in the Breakdown of the Theory of the Bicameral Mind

Filed under: Hobbyism and Nerdry,science!,thoughtful handwaving — Supersonic Man @ 11:15 am

I’m rereading the book The Origin of Consciousness in the Breakdown of the Bicameral Mind by Julian Jaynes. This is the kind of book that gets called seminal, because it is bursting with original thoughts and ideas and insights that can be quite inspiring to many readers. It is also the kind of book that gets called pseudoscientific crackpottery, because it’s also bursting with extravagant semi-untestable assertions based on thin evidence. I think many who’ve read it can relate to Richard Dawkins when he called it either complete rubbish or a work of consummate genius; nothing in between”, and even some of its defenders will nod their heads at statements like On first reading, [it] seemed one of the craziest books ever written” (Gregory Cochran). In modern terms, the book is usually described as “discredited”.

But despite the fact that I don’t think its controversial core theory is true, I still see a lot to value in the book, and recommend it as a thought-provoker. And one reason for this is because a lot of us still carry around a lot of naive ideas about conscious self-awareness that Jaynes does a good job of challenging, just by bringing an unusually clear eye to everyday acts of introspection.


November 9, 2021

the right wing’s rejection of reality is now its defining attribute

Filed under: Rantation and Politicizing,Uncategorized — Supersonic Man @ 11:17 pm

In earlier times, if we wanted to explain the difference between the left and the right in politics, we might talk about individual vs community, or big vs small government, or diplomacy vs militarism, or multiculturalism vs traditionalism, or equality vs inequality, or even communism vs capitalism. You could say it divides people between the compassionate and the selfish, or between the religious and the secular, or even just between the urban and the rural. But nowadays, in the United States of America, all such historical distinctions have now become secondary. What now separates our left and right is that one holds to reality while the other wholeheartedly rejects objective reality for fantasies and lies. One is sane and the other is not.

Some see this disconnection from objective reality as a sudden and startling transformation, but if you’ve been paying attention, it isn’t. Our right wing has been building up to this for a long time.


October 30, 2021

the novel “Dune” is both great and flawed

Filed under: Hobbyism and Nerdry,thoughtful handwaving — Supersonic Man @ 3:23 am

Dune, the career-defining 1965 novel by Frank Herbert which is the source of the current blockbuster film, is certainly a magnificent epic. It is a richly complex work with a lot of meaning, and it deconstructs a lot of narrative tropes familiar from our favorite legends and entertainment, at a time when such things were usually accepted uncritically. Hero’s journey, chosen one, white savior — they’re all there, played halfway straight but always askew and always illuminated from an unusual angle that makes you see it in a new way. And the book is packed with everything that was popular and trendy in its time, from martial arts to mind-expanding drugs, from youth rebellion to superpower origin stories, from the feudal struggles of high fantasy to the new science of ecology.

(Someone recently summarized the tale, when asked to boil it down to one sentence, as a bunch of greedy horny men failing to peacefully share a planet made of cocaine.)


September 27, 2021

the strange road to the Big Rock Candy Mountain

Filed under: Hobbyism and Nerdry — Supersonic Man @ 7:49 am

Some years ago, I looked out of an airplane into the southwestern desert and saw a remarkable sight: a mountain banded with an variety of rich colors. I pulled out my phone and managed to get a GPS fix, and was able to look up the location later. It turned out that what I had seen was the Big Rock Candy Mountain, which I had always assumed was an imaginary place.

postcard for Big Rock Candy Mountain, Utah

So I had to look up whether the song had been named after the mountain or the mountain had been named after the song. It was the latter: when the song became popular the residents near the mountain put up signs claiming the name for this volcanic hill, and also a sign proclaiming Lemonade Springs. And in learning this fact, I also learned something about the history of the song itself, which is a lot stranger than you’d think.


July 17, 2021

civilization is exploitation?

Filed under: Rantation and Politicizing — Supersonic Man @ 8:24 pm

In certain circles, it is not difficult to hear someone pronounce the phrase “Capitalism is exploitation.”  Is this valid?  I’m not gonna answer yes or no to that fraught question, as neither is entirely valid. What I am going to do is question the unstated implication that capitalism is somehow distinct in its exploitativity — the suggestion that since we’re equating exploitation to capitalism, this must mean that lots of other -isms are non-exploitative.  If you put capitalism up against all the other organizing systems it has competed with, such as monarchy and theocracy and feudalism and colonialism and warlordism and all the rest… well, it’s less exploitative than most.

The history of civilization mostly consists of an endless succession of different kinds of emperors, god-kings, priesthoods, titled nobilities, conquering hordes, colonial occupiers, and other groups which use some combination of social control and brute violence to put themselves in a privileged position above other people, where they get to keep the lion’s share of whatever they want.  And what they usually want in the largest quantity is labor.  Ever since the human race started to urbanize, the fabric of every kind of society that we call civilized has usually consisted of a set of rules which define one group of people which serve and another which are served.  In every land and in every age, self-appointed lords have declared themselves entitled to direct the labor of the masses, often for the benefit of their own class or their own family.  And fighting and dying for their rulers in warfare was just one more form of this labor.


July 15, 2021

a comparison guide to asshole space billionaires

Filed under: spaaaace! — Supersonic Man @ 9:26 pm

Elon Musk (SpaceX):

  • wealth comes from overvalued Tesla stock that he can’t sell without crashing the price, so it’s mostly illusory
  • forced the entire car industry to start shifting to electric motors, so he has probably done more than any other person to reduce global warming
  • approaches all problems by thinking first of basic physics, then engineering, then manufacturing at scale, then financial stuff last
  • the only private space entrepreneur to successfully sell satellite launches in quantity and make a profit at it
  • reusing boosters forced the other rocket builders to scramble just as hard as the auto companies
  • in that scramble, the ones who suffered the worst losses were the Russians
  • will soon send tourists on a multiday orbital flight, which makes the brief suborbital hops offered by the next two look pretty feeble
  • is now filling the sky with thousands of internet satellites which will make life tough for astronomers and might trigger a catastrophic space junk crisis (google “Kessler syndrome”)
  • wants to move a million people permanently to Mars
  • had a child with pop star and art dilettante Grimes; naming the kid “X Æ A-12” was apparently her idea
  • can’t keep his mouth shut when ordered by the SEC to stop tweeting things that influence stock prices
  • other mouthing off has gotten him in legal trouble… and he’s in court right now, arguing that it’s good for the company because being “entertaining” saves advertising costs
  • denied the severity of covid and tried to force all his workers to stay in factories at the height of the pandemic
  • generally overworks employees, and responds with threats if they mention unionizing
  • abusive tirades have been reported
  • has a hiring philosophy of “no assholes”, but is one

Jeff Bezos (Blue Origin):

  • wealth comes from paying people the lowest possible wages for the hardest possible warehouse labor, with intentionally high rates of burnout and turnover
  • Amazon makes very little profit, so all the money it makes somehow ends up owned by Bezos rather than the company
  • said to be intensely envious of Musk, trying to equal his accomplishments without realizing that you’d have to equal his brains first
  • a Trump-aligned tabloid once tried to blackmail him over infidelity, and found out that Jeff don’t blackmail
  • but what they found may have brought about the biggest divorce property division of all time
  • spent billions and billions on space for twenty years without launching a single thing for a paying customer
  • sold United Launch Alliance (the Pentagon’s favorite rocket builder) on a new engine for their forthcoming Vulcan rocket, and is now failing to deliver the engine, forcing the Vulcan to be delayed
  • also wants to fill the sky with thousands of internet satellites, but will probably fail to compete with SpaceX’s version, thereby rendering the satellite swarm useless as well as obtrusive
  • wants to move a million people to Earth orbit, along with heavy industry to relieve the environment down here
  • suckered some schmuck into bidding $26 million to sit next to him on the first brief tourist flight of the New Shepard
  • is a way bigger asshole than Musk… or at least he used to be

Richard Branson (Virgin Galactic, Virgin Orbit):

  • wealth comes from selling music, airline travel, railroads, hotels, gyms, clothing, advertising, cellphone service, motorbike taxis, books, business services… just about anything they could think of, all under one name
  • but for some of those he just sells companies the right to use the Virgin name, like Trump does
  • why just have one space company when you can have two?
  • unlike Bezos, has actually delivered a satellite into orbit, and people to zero gee (though Bezos is days away from catching up on the latter)
  • started accepting money for SpaceShipTwo tickets back in 2006, but the first flight with private passengers was not until 2021
  • but on the other hand, was offered a billion dollar investment by the Saudis and turned it down over their human rights problems
  • has no plans to move a million people anywhere
  • was an “adventurer” before he ever dabbled in space… crossed the Pacific in a hot air balloon, etc
  • his love life is apparently adventurous too, with at least one open marriage, and a tendency to inappropriate behavior when drinking
  • likes recreational drugs and wants them legalized
  • likes to come up with creative ways to cheat on taxes, even after once being jailed for it
  • has gotten four people killed working on his rockets, and four more severely injured, in two separate incidents
  • yeah, these things qualify him as an asshole

Max Polyakov? (Firefly Aerospace):

  • I can’t verify for sure whether he qualifies as a billionaire
  • wealth comes mainly from commercial real estate and e-commerce
  • grew up in Ukraine, where his parents worked in aerospace, and where he has founded a new engineering school
  • moved to Silicon Valley and then to Scotland
  • was a co-owner of dating websites accused of being scams; for this reason, subject to distrust in the business world
  • …on top of the distrust they already have for the idea of mixing American national security work with Ukrainian aerospace companies that may be influenced by gangsters
  • Polyakov’s main goal other than profit is probably to revitalize high-tech industry in his homeland, which has suffered many setbacks
  • Firefly was sued by Virgin Orbit and bankrupted by legal troubles; that’s when Polyakov bought it and revived it
  • Firefly built their first completed rocket about eight months ago and delivered it to Vandenberg, but we’re still waiting for them to attempt to launch it
  • Firefly’s only advantage in competing with Virgin Orbit or Rocket Lab is that they can (theoretically) lift somewhat heavier satellites, at a substantial cost increase
  • much less public than the others, so I have no solid data on to what degree he’s an asshole

In summary, Musk is the most innovative, Bezos is the most despicable, Branson is the most reckless, and Polyakov is the most unimportant.

I haven’t mentioned the tendency to overpromise and exaggerate what their companies will do, because all new-space rocket company owners do that to roughly equal degrees, whether they’re billionaires or not.

June 25, 2021

speciation of words

Filed under: Hobbyism and Nerdry — Supersonic Man @ 9:03 pm

When you look at the roots of our vocabulary, a surprisingly small set of root concepts gets elaborated into an amazing variety of particular terms.  By making little changes to an existing word you can split it into two distinct words, and by making metaphorical use of a word consistently enough, you can change its perceived meaning until its new form is heard as literal and concrete.  We don’t think of, for instance, the word “indented” as meaning like a toothmark, or “digital” as meaning like fingers. We don’t think of words like “inflamed” or “florid” or “petrified” as metaphorical, but they are, referencing fire, flowers, and rock respectively.

For an example of this divergence, consider the words “glazier” and “glacier”.  One is a trade profession and the other is a geological phenomenon, yet they differ in only one letter, and both have roots in the same glac- root word, which referred to ice in Roman times, and before that to anything shiny.  Glass, the substance, might well have been so named because it looked like ice to the first people who dealt with it… but apparently that’s not actually the case: the sense of “shiny” passed through Germanic languages into English without going through Latin and without specifically meaning ice.

Some other words in English coming from the same origin include “gloss” and “glaze” and “glare”, and medical terms like “glaucoma”… maybe even “glow”.  Our modern words for the sheen of varnish, the frosting on a donut, and a hostile stare may have originated with them being metaphorically compared to ice or glass.

For another example, consider words like “plan”, “plane”, and “plain”.  We use these for all sorts of meanings: a scheme or plot, a scale diagram, an abstract surface in geometry, a heavier-than-air flying vehicle, a tool for shaving wood, a meadow or prairie, and the quality of being bland and unadorned and ordinary.  We even use it for the name of a musical instrument, by borrowing the Italian version of the word, “piano”, which has almost all of that same range of meanings, from scheme to meadow, and also meanings such as floor or deck.  What do they have in common?  Every one of them got its name, directly or indirectly, from resemblance to a flat surface.  The sense of scheme or plot apparently derived from the sense of, like, “floor plan” — a 2D diagram for how you intend to build something. (And there are plenty of additional similar words related to flatness from the same or similar roots, such as “plank” and “flake” and “placid” and “flag”.)

There’s nothing flat about a piano, but in the Italian musical tradition they started using the word “piano” in the sense of bland ordinariness to mean playing softly, and then they combined it with “forte” (strong) to make a portmanteau word to describe a keyboard instrument that could play soft or loud notes at will, and people shortened it to piano, keeping just the part of the name that described its ability to play some notes without emphasis — you might say, flatly.

It’s pretty common for a foreign word such as piano to be borrowed for a narrow specific meaning when, in its original language, it had a much broader and more general range of meaning.  For instance, in English we use “chef” to mean a haute-cuisine cook, but in French the word just means any kind of chief or boss or leader.  The German word “gestalt” basically just means “shape”, but in English we use it only in particular abstract and abstruse ways.  In Japanese the word “sensei” means any kind of teacher, and “bukkake” just means “splash”.  “Kung fu” (or gōngfu in modern pinyin) means mastery or accomplishment in general, without referring to any particular area of skill.

The process goes the other way as well.  “Handy” is a fairly broad word in English  but when borrowed in German it means cell phone.  “Smoking” in French or Spanish means a formal jacket or tuxedo, and “pull” in French means a pullover sweater or jersey.  Words like “surf” or “piercing” have many senses in English but usually only one when borrowed into a language like Spanish.  In Japanese, the borrowed German word “arbeit” (or arubaito), means specifically a part-time job rather than work in general, “bike” (baiku) means motorcycle but not bicycle, and “charge” (chāji) means making an electronic payment, and only in some contexts.

One notorious example of divergence is the set of words arising from a root of blu- or ble-, which referred to the sky, from an earlier bhel- that indicated brightness.  This gave us not only the color “blue”, but a whole range of words related to paleness or featurelessness including “blank”, “blanket”, “blanch”, “bleak”, “bleach”, “blush”, “bland”, “blonde”, “blend”, “blink”, and “blind”.  In romance languages it became blanc, meaning white… yet by another path some say it is also the origin of “black”.  The closely related alb- root, meaning “white”, has plenty more derivatives.

Words from “edge” to “aggressive” to “acumen” to “acid” to “acne” to “exacerbate” to “mediochre” to “paragon” to “oxygen” all come from a common ak- root which means sharp or pointy.  Even the word “hammer” shares this origin according to some, though all resemblance is now lost.

In these various kinds of mutations, at least the spelling and pronunciation often change to accommodate the added meaning.  We can easily hear the difference between, for instance, “extract” and “traction” and “tractate”.  But what about the simple word “track”?  In the dictionary I see about three dozen distinct meanings, from the act of monitoring a flying object’s location by radar to the linked steel belt that a bulldozer rides on, all using exactly the same spelling and sound, and all derived in some metaphorical way from the concept of a footpath, or footprints.

And there’s another gotcha here… some of the words I mentioned earlier, and related ones like “tractable” and “tractor” and “tract”, are apparently derived from a separate root unrelated to “track” — maybe even more than one other root.  And “trachea” is from yet another distinct root, though it sounds like it could plausibly have come from a word for a path.  Sometimes as words diverge from each other and spread to cover a wider area in the space of pronounceable sounds, they run into other words traveling in the opposite direction… for example, the noun “egg” (what birds lay) has a different root from the verb “egg” (to provoke or goad), with the latter being another derivative of ak-.  They’re probably no more connected than the Japanese car make Mazda (a flattened form of Matsuda) is to the Zoroastrian deity Ahura Mazda (since flattened to Ormazd)… but it’s hard to be certain. Some of these may be words that started together and drifted apart in ancient unknowable times before swinging toward each other again, while others may have begun farther apart than we’ll ever know. However far apart they began, often when they get close together people will just start pronouncing and spelling it the same as a familiar word, snapping them together. For instance, the Germanic word “schrubben” (to clean with a brush) may have become “scrub” in English because that was already a familiar word for bushes and small trees.

I still prefer this kind of back-and-forth spread over the situation where more and more meanings are piled onto a word without it changing.  The worst case I can think of is the word “set”.  One dictionary I checked listed 93 distinct meanings for this word!  It does seem like most of them derive in some way from the concept of remaining in a particular place, though that’s not very apparent in cases like the setting of the sun, or a set of golf clubs.  And the words “sit” and “site” and “settle” and “sediment” and “situation” clearly derive from that same sense of fixed location… so it could have been worse.

Personally, I’m glad of the variety — I appreciate the value of having words drift far enough apart so that the musical term “opera” and the mathematical term “operand” are not easily confused with each other.  (Both derive from a root that means work or effort.)  And it’s impressive what a range of variety and specificity we’ve come up with, starting from a far more limited range of simple words for things like stone and sky and wind and water.

In this, evolution of words really does resemble the evolution of life.  Seeing all these root word relationships, it would be pretty much impossible to argue, say, that language was created whole and complete in one act.  It’s clearly grown over the years to be much larger and wider than it once was.  But the diverging nature of life is nearly as self-evident, and people deny that all the time.

Next Page »

Blog at