(From Lognet 97/2. Used with the permission of The Loglan Institute, Inc.)

Lo Lerci

(Letters)


This first letter is from Mark Zacharias:

Hoi Djimbraon, fie:

Don’t feel bad about taking so long to answer my e-mail. ... I know from reading Lognet how busy you have been.

... I continue to enjoy Lognet; every issue seems to cover fascinating new ground. The various pieces on the ontological status of masses, sets, etc. have kept me thinking for some time. The first issue of La Logli is still in my “to be read” queue.

As for the language itself, I continue to use bits of Loglan for various purposes, such as for minor notes to myself, as a unique greeting (Loi, tu he[?]), or as a substitute numbering system (i.e., ne, to, te, etc.) in place of Roman numerals in outlines. My expertise in Loglan is holding steady. Someday when I have more time to practice I hope to improve on it.

Thank you for your service.

Hue Markzakar[a]iys

I am beginning to think that people can become active logli only when they retire, soi crano. Loglanizing seems to be hard to confine to weekends. I know! I try to write books on other (utopian) topics in addition to doing my share of loglan work, and I find it’s hard to do both.

It’s always very satisfying, however, to learn that our efforts, in this tiny little non-profit publishing house we call “The Loglan Institute”, are giving both enjoyment of, and instruction in, this endlessly fascinating little language, both to our readers and to ourselves. Keep us “reinforced”, as the Skinnerians say...or used to say; there may not be many of them left! But little notes of appreciation of this kind always make our day.

This next letter was e-mailed to me by Tiftua Emerson Mitchell in November 1996 in answer to some questions about the formal (i.e., transformational) relationship between me and lo that I had put to our lodtua munce. E’s response was so completely clarifying, and thus, it seemed to me, potentially so useful to other logli that I asked E for permission to publish it (which E always gives, of course, soi crano, being unwilling to hide his light under a bushel).

Hoi Fremi:

You ask [after specifying some reasonable assumptions about their usages:], “Would we not be implying that me and lo are inverses of one another, wherever they occur?”

Your series of equivalents

  a) Ta meDNA.

b) Ta melo deoksiribonukle acdi.

c) Ta deoksiribonukle acdi.

certainly would show that melo cancels. However, a true inverse would cancel the other way around also: lo me [would have to drop out, too].

I do not think this other cancellation should happen. The reason is that there is only one “predifier” (me), but there are several articles that make [“predifiable”] terms: le, lo, la ...

It seems to me that logically mela, mele, melo, etc. would all cancel. The distinctions among le, la, lo, etc. are not part of the predicate, and hence are irrelevant to the meaning of mel_ preda == preda. 

The other way around, le me preda, etc., has the distinctions put back in. The meaning of the following construction ought to be clear:

Le melo deoksiribonukle acdi.

 (it parses! I am getting better at this, slowly.) And [its designatum is] clearly different from [that of]:

Lo deoksiribonukle acdi.

Bottom line: I think that all the mel_ should cancel out, that melo ought to, and that l_ me should not.

Technically this would make me a “left inverse” of the l_ articles, but not a true inverse.

—Hue la Logli E’mrsn

Thank you, E’mrsn; this does it for me.

The following letter is made from two letters received from Steve Lytle, a new tiftua whom we’re finding a bit hard to put to work. The reason is that, unlike our other tiftua, S is not on e-mail! So with S we can’t do those instant communications, those transfers of long work-files that seem to be so productive of cooperative work. But obviously we’re determined to solve this communication problem with S...even if it does force us, soi crano, into the antique posture of writing something on paper from time to time!


Dear Jim:

I have an IBM clone, i.e., a DOS machine ... and some word processors that do not include MS Word [the processor used by most of us who work for TLI].

Unfortunately, I suspect my 3.5-inch disk drive is faulty, because I can’t use the software you sent me. It (the software) does work on my father-in-law’s computer (actually, MacTeach does; I still haven’t gotten ... LOD to work, so I don’t think it’s the disk. I’m planning to get it (the drive) repaired soon. [I believe, on the strength of that, I sent Steve a new LOD; but if I didn’t, Steve, please remind me to do so if this important tool still isn’t working for you when you’ve repaired your drive.]

... Are there version numbers or some similar way to differentiate the stages of development Loglan has gone through? If so, what is the current one?


This is one of those provocative questions that are hard for me to resist. I know so much more about the long history of this project than anybody else, that I feel a moral obligation to share my knowledge with anyone who asks for it. So I’ve made the following long, dutiful answer to S’s question. Sorry if I repeat some things that some others of you know.

The history of LIP, our “Loglan Interactive Parser”, probably provides us with the best tracking of the development of the language over the years, at least of its grammatical structure. Unfortunately LOD (the “Loglan Online Dictionary”) does not record lexical development of L in quite the same way, though its dated entries do provide some clues as to how it grew. Still, “grammar” is broadly enough conceived by LIP so that it reflects structural changes in the lexicon—the addition of new lexemes, etc.—as well as in the grammar rules.

The current version number of LIP is 2.24; and the grammar with which it is currently loaded is Grammar No. 81. That last number means that, since early 1982, when the first Loglan grammar that “parsed a Corpus” was discovered by Scott Layson and me, there have been 80 subsequent grammars, each one incorporating one or more new, if sometimes subtle, changes in the formal structure of the language.

To keep up with all this growth—to keep our grammar unambiguous at the same time that it was being pummeled by all those “growth hormones”—the Test Corpus that we used to test the conflict-freeness of any new state of the grammar had also to be constantly expanded. The Corpus itself—which is in the keeping of Dr. Robert McIvor, our Takrultua, and is the main archival record of what the language has ever accepted as grammatical—thus reflects the increasing scope and power of the language.

Happily enough, grammatical change is also turning out to be a slowing process. That is, there was less of it last year than there was the year before, and so on back to 1982 when it all began. But change is also an apparently unending process. There is no sign yet of its ever stopping, and many signs that logli, being logli, will be forever finding new and sharper ways of saying things logically. It is therefore a pretty sound bet that, if L survives at all, changes in its grammar will continue to happen for a very long time...but that its grammar will also stay unambiguous for all that time.

For the truly amazing thing about this process is that, for fifteen years now, we have managed to accommodate all the changes that logli have come up with, but in ways that have kept the grammar continuously unambiguous. Our takrultua, Bob McIvor, armed as he is with the powerful diagnostic tools developed in 1978-1982, and described in La Logli 96/1 (“An Unambiguous Grammar for Loglan”), has had no trouble at all in keeping up with the changes that the Keugru keeps shooting at him. (The magnitude of the Keugru output—the K having turned out to be not just the monitor of change in Loglan but also its most abundant generator—is suggested by an article that also appeared in that first issue of LL, namely the one called “Loglan 1 Updater”. This piece by Kirk Sattley reports, in calendrical order, all the grammatical and other structural changes that have taken place in L in the past eight years....since the publication of the 4th Edition of Loglan 1 in 1989, in fact.)

Bob is now increasing the scope of our grammar-management tools. For they will soon include software that will not only parse Loglan sentences but also write correct Loglan text from input speech. The new software will also produce correct Loglan speech from input text when we provide it with the tools of speech; and that means that, used as a learning tool by darli logli, it will also be able to correct both the writing and the speech of human users as well as their grammar. In short, the new software should enable this expanded LIP to function as a gentle teacher of everything we do in L except think. (Now, who would like to get to work on that addition to LIP?)

These new tools—Bob is calling them the “Resolver/Producer” set—will actually “produce” L speech from L text (i.e., read it aloud) and “resolve” the flow of any L speech it hears into a fully punctuated stream of L text. Thus, it will teach us to hear L speech correctly—as measured by our ability to transcribe it into text—and, in its other mode, it will teach us to speak correctly, as measured by the ability of others to transcribe accurately what we’ve said.

In other words, a Resolver/Producer project is now underway at The Institute that will hugely increase the scope of the computer-aided tools with which we can correct and improve our own use of the language...although still living, as we do, at considerable distances from one another. There is more in this issue about the Resolver/Producer project.

... Lemi komta na djela. I ao mi helba tu. [Steve is telling me that his computer is now well, and that he wants to help “me”. If tuu—the new personal pronoun which designates the set composed of the auditor/reader and some unspecified other(s)—had been available, I imagine S would have used it instead of tu.]

Now for some of S’s comments and questions:

 Has anyone tried to teach Loglan to an infant as its first language, or concurrently with a first language? After all, teaching is a great way to learn, and having a native speaker would be a huge boon to the language.

I agree it is and would. But to the best of my knowledge this has not happened yet...though several junti logli tora (young loglanist couples) of my acquaintance have threatened, soi crano, to do so. What happened in each case, I think, is that the parents did not think either themselves sufficiently competent L-speakers to serve as role-models for their children or L sufficiently settled to be taught to an infant. Still, as you say, attempting to teach something is probably the best way of learning it. So one would think that if L were settled enough to justify the undertaking, then keeping abreast of a child’s linguistic development would be a pleasant sort of challenge rather than a forbidding one…provided, as I say, the language itself had reached a point where one could count on its basics, in any case, remaining unchanged. It may be that, with the newest additions to its personal pronoun list—see the inside backcover of LL 97/1 for a table of the new plural pronouns—Loglan may finally have reached that point! Inumoi nu hapvia, Hoi Logli Penre! I ea muu madzo su brana logli vi levi telfoa! (So be welcomed (happily-seen), O Logli Parent! And let’s (you and I and others collectively) make some (one or more) born logli on this planet (Earth-form).

 

I would prefer L3 as a notebook filler. (See the last sentence on p.27, Lognet 96/1.) I have bound my copy of NB3 and am very pleased with it. [The production of notebook fillers should be quite easy once the L3 copy is camera-ready for publication in LL. Every 8.5” x 11” page, the size of the usual notebook page, contains exactly two 5.5” x 8.5” pages, the size of LL pages. I expect we could ask Pagemaker to print it out that way as well. At first we could make copies of the Notebook filler available only by special order; later, perhaps, if there’s enough interest in it, we could invest in a printing run.]

 Lately there has been some experimentation with the “punctuated style” and leaving out pause-commas that are misleading or offensive to the eye. As a beginner, I like to see all the words written out, including the quotes (“ and ”) and [letters used as] anaphora. [As a beginner, indeed you should. The first six lessons of Loglan 3 were entirely “phonemic” in that sense, i.e., every L-word in them was spelled out in letters used phonemically. Starting in Lesson 7, however, which is the first lesson in the upcoming Vol. 2, we’ve experimented with leaving out the post-name pause-commas and “punctuating” a few other types of words, for example, some letter-variables. But we plan to play this by ear in the later lessons, listening very closely to what our cirna have to say! This is the great advantage of publishing L3 as a serial first, of course.]

 A larger viewpoint of the preceding comment shows that all capitalization is unnecessary. If the spoken and written forms of Loglan are supposed to [be] isomorphic (and I think they should [be]), then all the words could be written out without capitalization [or without non-obligatory interverbal spaces, for that matter!]. On the other hand, the numerals (digits) are useful symbols. The capital letters could be used as anaphora.


In fact there has been a long and interesting behind-the-scenes discussion of this very issue. The “isomorphic purists”, such as yourself, have wanted to go in one direction while the “isomorphic pragmatists”, of whom I count at least my recent self one (though I was originally a purist), have wanted to go in another. In recent decades the pragmatists seem to have prevailed. No one wants to give up the isomorphic principle, of course. We all agree that how an utterance is spoken in Loglan must be capable of specifying within very narrow limits how it is to be written; and vice versa. This has always been agreed to by everyone. That is, there are no “anti-isomorphicists” among us.

The purists, however, have frequently wanted the structure of speech to be approached in text as closely as possible. (Even purists, however, draw the line at leaving out interverbal spaces. That is, they don’t want us to write lemiko’mtanadje’la, although that seemingly “one-word utterance” would resolve into four “real” words perfectly well. (Did you know that ancient Greek was originally written in this way?) You almost have to read that textual expression aloud to “get” it, however! What it says is Lemi komta na djela = My computer is now well.) But most purists do disapprove, as you do, of using capitalized initials on names and utterances.

The pragmatists—of whom I’ve confessed to being one...and I admit that it’s a little unfair that I am, as founders have a certain honorary weight in such discussions, soi crano—have recently been focussing on the problems encountered by the reading or text-searching eye. Whatever conventions used in natural language texts that demonstrably help the eye read a block of text quickly, or scan it accurately for specific content, should, we pragmatists feel, be adopted for Loglan as well. Thus first-letter capitalization of names and utterances was reinstalled in the language around 1962 because it demonstrably aided readers to do just that. In the earliest L texts—those of 1955-60, when even I was a purist (you will be interested to learn, or may even have noticed, from the SA article)—capitals were not used. But it soon became apparent, as I and my University of Florida colleagues on the NIH-supported Loglan Project quickly discovered, that initial capitals simply made it easier to find such important objects as names and utterance-heads in the flow of text.

In a similar way, using numerals and letterals instead of numberwords and letterwords in certain kinds of texts—writing 2 and t instead of to and tei, for example—also helps the reader in r’s task by making the text so-treated more compact and its symbolic structure plainer. Also, if somewhat paradoxically, with these well-known but non-phonemic objects scattered through it, such text is actually easier to read. “Punctuation”, in the sense of writing « and » for li and lu, and  ( and ) for kie and kiu, has a similarly soothing effect. For this practice, too, helps the knowing eye read swiftly, both by compacting text and by making the visual signs of these common objects stand out in the field of letters.

But, as you see, we’re still experimenting with this reader-facilitating project...especially now that the Resolver Project has been resumed. We all agree with you, however, that, for early learners, unpunctuated strings of phonemically-used letters is still the way to go. Even for such learners, however, the initial letters of names and utterances should, we believe, continue to be capitalized. The learner, too, needs to locate ellself from time to time. (There. You see? I’ve spelled the letterword ell out for you, soi clafo!)

 

Regarding the number system: I haven’t seen this issue raised at all, but it annoys me all the same. I’m referring to the current practice of naming numbers. If you compare numbers with words, then the meaning of the word is lost when the word is spelled out. The “meaning” of a number is the digits used to make it up plus an indication of the size of the number (i.e., its magnitude) as given by place value. These place values are what is missing in Loglan.

I once invented a system of naming whole numbers that minimized the size of the words. Each digit 0-9 was represented by a consonant letter and the place value (the power of ten) was indicated by a vowel letter. If m stands for 1 and d for 2; i for 0, a for 10, and u for 100; then 12 would be madi and 200 would be du. The vowel e was for thousands, and o was for the next higher unit when all the lower number names were exhausted. The whole table of values I used was:

1 m 6 s n*1 i

2 d 7 p n*10 a

3 t 8 v n*100 u

4 k 9 n 1000n e

5 c 0 z 1030n o

[Examples:]

1997 = mime nu na pi

  = one thousand nine hundred ninety seven

5,000,000 = cide

  =five one-thousand-squared

Hue Stiv

Admirable for large numbers but hardly “minimal” for smallish or middle-sized ones, don’t you think? And the latter are much more common in speech. Thus the standard L word for 1997, nevevese, has a length of 8 phonemes compared to your word mimenunapi with its 10. And all that is required to understand the L word is a positional interpretation of the “place values” of the digits, something that is understood by all peoples who use the decimal system...now practically universal. Also, is your place-value system really any better than the floating-point one now in such common scientific use? In the latter system, 5,000,000 = 5×106, which is often further abbreviated to the 3-character expression 5E6. Five-ee-six admittedly takes longer to say than cide, but not much longer. We don’t have a symbol for E (read ...times 10 to exponent...) in L yet, but we could easily adopt one. There’s no reason why we, too, couldn’t use Ema in such expressions, and symbolize it with E following the current international custom.

Then the Loglan floating point representation of     five million would be feEmaso (5E6), pronounced                       /fe,eMAso/, which uses 7 phonemes...but only 3 characters compared to cide’s 4. But curiously enough, femomo (5θθ), with its use of our “triple zero” (θ), is even shorter in speech, using only 6 phonemes, although with the same length (3 characters) in text. Plunging ahead on this course, a googol could then be either neEmanema = 1E1Ø, at 9 phonemes and 4 characters, or nenisuanema = 101Ø, at 11 and 5 (Ø is the L “double zero” ma, and ↑, the sign of exponentiation (“to the power of”) sua), while a googolplex could be as short as 14 phonemes in speech or 6 characters in text: neEmaneEmanema = 1E1E1Ø, or, only a little longer, but perhaps more intelligibly, as neniEmaneniEmanema = 10101Ø, which, except for our exponentiation arrows (which linearize these expressions) and our double-zero (which shortens them), is the internationally standard way of writing this large number.

I have no doubt that Steve’s system would produce words for these very large numbers that are shorter still; but I’ll make no attempt to produce them here as I don’t understand S’s translation of his suffix -o as 1030n. I do wonder, however, if whatever economy S’s system achieves would be worth the cost of learning an entirely new “place-value” system instead of the simple, and already widely-known, extension of the international  decimal system that the floating-point system represents.

* * *

Reed Riner, the Northern Arizona University anthropologist who wrote “Loglan and the Option of Clarity” for Et Cetera in 1990, tells me that he still gets “1+ request a month” for reprints of this article. Reed recently forwarded me a thank-you note that he’d received from one of the most recent recipients of this excellent paper. (Write either me or R if you don’t have “Option” and would like to see it.) The recipient’s note was sent Reed by e-mail, and written with that admirable brevity that e-mail seems to stimulate. It had an interesting final twist that I’m sure you’ll appreciate.

[Dear Reed:].

I think that I never thanked you for [the reprint of your “Loglan and the Option of Clarity”, for which [omission] I heartily apologize.

It is a fascinating article, for which I heartily thank you.

You will see below that I am a lojban supporter, but by chance rather than by design. It’s all much-of-a-muchness to me, seeing that they have concentric aims.

Thanks again,

co’i mi’e nikos

[This is apparently Lojban for Hue Niklys (Symrz)]

It’s both interesting and pleasing to me—as the reluctant father of one of these languages and the proud one of the other—to observe that at least some people are now sorting themselves out pretty much by accident into the two “logical language” camps...that the angry (and rather foolish) side-taking of the late ’80s and early ’90s seems to be about over. Still, although we and the lojbanists may have “concentric aims” (I doubt that we still do, but that’s an intriguing metaphor, soi crano, causing me to wonder, Are our aims a subset of theirs? Or theirs a subset of ours?), at least the products of all that energetic aiming are still joyously dissimilar. Loglan is still the mother-language, the living beauty to all who know her, turning into something even closer to a natural language everyday. I’m just a little prejudiced in favor of my “legitimate” offspring, of course, as fathers tend to be. For she is the one on which I’ve been able to lavish all this care.

* * *

This next letter is from Herschel Elliott, one of my oldest friends in both senses of the word old. H is now a “retired” philosopher (but do philosophers ever retire?) and one of our Trustees. H is also one of the earliest contributors to Loglan, being the author of L’s “implicit quantification” system. H originally invented this system to use in teaching symbolic logic, and it worked so well for him in the classroom that H suggested that I incorporate it into L. As implicit quantification seemed to fit the L project of “making logic easy”, I agreed, and H helped me build it into L during the Fall of 1973, when we happened both to be in Gainesville. Implicit quantification was one of the last modifications made to the pre-1975 language, and hence to the 2nd Edition of Loglan 1, then being revised to produce the 3rd...which is the one that got published in paperback in 1975 and brought so many of you aboard. 

What follows is a Loglan-relevant excerpt from a very much longer letter from H to me and my wife Evy. In the passage I’ve selected for reproduction here, H is raising a delightful quibble with me about whether “sin” can be defined in L. At least it’s the sort of quibble that people with loglafied heads tend to find delightful, soi clafo.

Dear Jim and Evy:

... The July 1996 La Logli is a nice-looking, good little journal. Congratulations! [The congratulations  should go to tiftua Alex Leith and Kirk Sattley, who put it all together. So in their names, Nemosia, Hoi Hrc! (A thousand thanks, O Hersch!)]

I note on page 32 at the bottom: Lopo nuo mormao ga po lidzao = Suicide is a sin. But isn’t translating the [E-]word sin into Loglan to incorporate metaphysics into the language? [Yup, I’m afraid it is.]

The other night I heard an interesting discussion on the BBC. It was discussing English as a world language and the difficulties of translation [that would result]. A discussant mentioned, for example, that it was impossible to translate the English word sin into Chinese. There was a word for crime in Chinese. But there is no way to get the concept of sin across.


I doubt that. Given the Chinese talent for metaphor—which must include a talent for understanding metaphor—there will always be a Chinese way, through metaphor, for understanding those alien corners of their world into which their current vocabulary doesn’t reach. For example, how about our telling the Chinese that the English word sin predicates something that is “religiously bad”? C has words for both bad (hwai) and religion (dzungjyau); so a metaphor for sin, dzungjyau hwai (please forgive my bad Chinese, Hoi Jungi!), comes readily to mind. Moreover, it is likely to be at least approximately understood by any Chinese mind.

But, we must object for H, since the word sin is an English word, it must mean bad as assessed by some English religious canon. Well, the Chinese have a word for English, too, namely yingwen; so the metaphor we finally require to explain the notion of English sin is a three-termer: yingwen dzungjyau hwai. (The modification order is the same in C, namely AN, as it is in L.)

In any case, “bad in the light of some specific religion” is what the L-word lidzao will mean to the knowing logli, who might one day be a loglentaa jungi (a Loglan-speaking Chinese person) as well. And j (that jungi) will then be able to make these inferences from the metaphor underlying lidzao just as you and I can. For its deriving metaphor—as will always be known to logli—is nothing more nor less than lidji zavlo = religious(ly)-bad. 

Finally we must anticipate H’s objection: But does lidzao translate the E-word sin exactly? Not exactly. Lidzao is a much wiser, more understanding, less provincial word than the word sin is; that is, we can predict that to some miniscule but measurable extent lidzao will promote the qualities of wisdom, tolerance, and cross-cultural understanding in its users as use of the E-word sin does not. For lidzao does convey the idea of sin admirably. What it suggests is that there are some religions that define some of the behavior of at least some of their adherents as bad. And that is true. The word for that behavior in E is sin and in L it is lidzao, which is thus, by comparison, an understanding-promoting, anthropologically rather sophisticated word. In short, a good metaphor should be more than a mnemonic; it should also provide a semantic route for understanding strange concepts, and this one does. (Note that sin, being primitive in E, provides no such route. Like all primitives, it is a take-it-or-leave-it word.) H now continues.

Sin is a monotheistic metaphysical concept. As Jesus said, if a man looketh upon a woman with lust, he has committed a sin even if he did nothing. Sin is mental turpitude. We are all born in sin, even though we did nothing but get born. And the wages of sin is eternal damnation. Come to Jesus and be saved. By his willing[ness] to suffer death by crucificion, he will wipe your sins away. His eternal love will force all who do not accept this line of [reasoning] into eternal torture. Arrogant, cruel, vindictive [person] that He [evidently] is.

I have softened some of H’s language. But this is Christian sin, surely. In other religions, quite different sets of behaviors are held to be bad. Among the Jains of India, for example, stepping on ants is bad. As Jainism is a religion, that makes ant-stepping-upon an instance of lidzao, if not of Christian sin! (By the way, I have just discovered that lidzao was not defined in this religion-specific way in LOD! It is now; I’ve just redefined it, soi crano. (Now that LOD is electronic, corrections in it can be made by anyone, and very rapidly. I hope other users of LOD are doing so. Just send your corrections in to Bob McIvor, who is our Purbuktua as well as our Takrultua. Bob wears many hats.) H continues.

Needless to say that Chinese [persons] cannot understand “sin” and cannot translate it. [Possibly. But lidzao says something that Chinese logli will be able to understand. So, when a loglentaa jungi reads a translation of some E text into L, j may come to understand lopo kristno lidzao (Christian sin) as well, partly from the way the concept is expressed in L. If j is also a translator of E texts into C, then simply knowing the L predicate for it (kristno lidzao) will no doubt help j translate the English, i.e., Christian, word sin into some suitable C metaphor...even if the one I’ve suggested above turns out not to be a very good one.]

Doubtless Loglan must be able to speak about sin and about phlogiston, angels, and jinni. But is it possible to differentiate between legitimate concepts and deceptive metaphysical concepts? Perhaps that problem cannot be solved by language. [It can be helped by clear metaphors, as I’ve tried to indicate above; but I don’t believe it can be formally solved.] But in that case, no language can be metaphysics-free. [That is true. But one can be metaphysics-minimal, which is all that L purports to be.] That distinction is one that must be made empirically, pragmatically, scientifically. 

.... My view is that scientific truth is just truth according to the presently accepted model or [its] paradigmatic assumptions, i.e., metaphysics. For example, is it true that the laws of matter and space are the same now as they were 10 billion years ago, the same here as everywhere else in the universe? Can such questions be solved empirically? I believe not. They are metaphysical. But they are at the core of science. ...

Love to you both,

Hersch


Herschel’s final paragraph moves me also to philosophize. If H means by “the laws of matter and space”just  their formulation in physics textbooks and similar venues, then obviously his first question, the one about the constancy and universality of scientific law, has an answer; and that answer is a resounding no. For scientific laws have changed fundamentally in every century since the scientific enterprise began. Moreover, we have found that out empirically, that is, by looking at the written records of science. This answer seems to serve for H’s second question, too: Are the metaphysical assumptions of science empirically investigatable? The answer is clearly yes; for the changing nature of both these laws and these assumptions has been found out by factual studies of our history, our biology, and our many human cultures. From these studies we have learned, first, that the “laws of nature” that we say “science has discovered” are human artifacts. They are like pots; they were made by human hands. They were not just found lying around in nature for us to pick up. Second, these laws went from non-existence to existence—a pretty massive “change”, that!—in, or in the centuries just before, the 7th Century B.C., and they have been changing substantially, fundamentally, and sometimes even startlingly (as during the Einsteinian paradigm-shift that occurred early in this century) during nearly every subsequent decade of scientific activity. The same is true of what we call the “scientific method”; for it, too, has undergone dramatic and fundamental changes since the beginning of science, changes often brought about through critical inspection by philosophers—sometimes by scientists themselves—of the assumptions scientists were making. So the answer is a three-fold yes: scientific laws, methods, and “metaphysical assumptions”all change in time, as what seems to be the natural outcome of the process of inquiry. 

Of course if H means by “the laws of matter and space” the features of the non-linguistic world to which these linguistically-stated and humanly-fabricated laws of ours refer, that is, the regularities of nature that are only partially, because always incompletely, described by these laws, then we do not know enough—yet—about the stability/instability of these extrahuman physical arrangements to know whether they have, or have not, changed—and certainly not whether they will or will not change—given suitably long passages of time. But clearly asking whether they have changed or not, or will change or not, are also empirical questions about the world, not about the symbolic games we play with one another in our attempts to understand the world. So one day we may have an answer to H’s question...even so-interpreted.

At the moment, however, we do not. It is interesting but not particularly relevant to that question that most physicists currently assume a negative answer to it: that these fundamental regularities in nature do not, and will not, change. But this can hardly be more than a convenient assumption, a methodological ploy; and as such, it may well turn out to be false. Such assumptions, these working principles of ours—the things we take for granted about the world during the course of our inquiries into it...the things “we leave unexamined about the process of inquiry while we’re using them to examine the world”, as John Dewey called the apparently circular process of inquiry—these are what are “at the core of science”. But they’re an impermanent lot, this core. They, too, change with the passage of time, with the increasing deftness of inquiry.

The fact that we don’t examine our most deeply buried methodological assumptions—at least not very often—doesn’t mean we can’t examine them. We sometimes do, and have. In fact, the greatest advances in philosophy—also its greatest liberations of the human spirit—seem to have come when those impudent persons whom we call philosophers—people like Socrates and Hume and my friend Hersch—started asking just those unasked, infuriating questions (“What do you know?”, “How do you know it?”, “How can you be sure?”, “How can you justify the principle of induction without using it?”, and “What justifies your using it if you can’t”?) that these people called philosophers have, by this time, got us used to. These are the “metaphysical”, the “epistemological”, and the “ontological” questions about which we must assume some answers in order to even start our inquiries.

As John Dewey, possibly the greatest student (so far) of human inquiry, pointed out some years ago—in a book called Logic: The Theory of Inquiry,  published in 1938, almost 60 years ago—even these answers, the ones we seem almost “obliged to give”before we can start thinking, also change with the passage of time. What seems to be “circular reasoning” (about induction, for example) is not; it’s a spiral. It looks like a circle when it’s viewed ahistorically, that is, in two dimensions; but it’s a circle that has been broken and stretched into a spiral by being laid out in time—and thus made three-dimensional—by the process of inquiry itself.

—Hue Djimbraon