DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT SOFTWARE OWNERSHIP STRUGGLES An Anthropological Approach v.0.1 Michael Turner 15 Feb 1999 ABSTRACT The central hypothesis of this essay is that all these struggles over Open Source definitions are rooted in human instincts about valuable information. The main notion to be defended in support of this hypothesis is that there can be "information instincts" in the first place. I argue that there are, and that they emerged naturally out of human attributes--natural language, which one observer, Steven Pinker, has boldly characterized as an instinct, together with the mobility that comes with the human ability to adapt to a huge range of ecosystems. Human beings make decisions about whether to hold information, whether to share it, and how to charge for it, based on in-born feelings as much as on economic rationality, and tension between the two engenders a conflict that will almost certainly be perpetual. SPOKEN LANGUAGE - THE ORIGINAL PROGRAMMING TECHNOLOGY If humans have instincts about information, most of them are probably observable in the medium that all cultures share, regardless of technological level. This is natural language So let's look at that. Human language ability appears to be hardwired to a great degree, but with a lot of what certain linguists (Chomskyans, notably) would call "parameters". Some of these parameters have dramatic effects. The grammar of some languages (Japanese, Turkish, Finnish, etc.) is almost totally the reverse of that of English or French. As "parameter settings", these orderings are pretty strong--try to speak a sentence backwards (even if you know both English and Japanese to native-level proficiency, say), and you'll see. Language seems to be defined to a great degree by these parameter settings plus vocabulary. All else is detail (though there's a lot of detail--lexicalists would say it's ALL detail.) Why bother with parameters? Why couldn't evolution have wired us all for one basic universal grammar that doesn't squirm around--and even vary rather radically--from one locale or tribe to another? Think of the trouble it would save. One possible answer might be genetic: evolution was on its way to this supposedly-optimal goal, but was at different points of fixed/floating parameters-setting combinations in different races before the various human strains went into a population explosion and melted into each other. This can't be right. If this were true, there'd be genetic propensities toward one family of grammars over another among races. Australian aborigines, in particular, would have difficulty learning and speaking English, considering how long they were isolated and how recently (and how little) they've interbred with non-native Australians. The fact is a native speaker is ANYONE of ANY race who happens to be born into a given speech community. Or communitIES--there are fully multilingual people, who are fluent native-level speakers of two or more radically different languages, and who can switch between them without taking a breath. But how many people have acquired that level of ability with more than one language in anything other than a full-fledged speech community setting, with full-fledged speech community participation? Such people are vanishingly rare. Worse, if you isolate two groups of native speakers, in a few generations, they'll start having a hard time understanding each other. If anything, evolution seems to have favored human language divergence, when you'd think that human language unification would have survival value at the individual, group AND species levels. So I guess language unification must not have value at all those levels. At some level, there's antagonism. I'd say it's the group level. If it was the individual level we'd all tend to speak what linguists call "idiolects" (dialects of one) far more than we do. And it's hard to imagine that a tendency toward grammatical unity wouldn't have species survival value. Given other human attributes (i.e., high adaptibility, and having language in the first place), the universal-non-parametric grammar's survival value was probably second-order. Only at the group level does the ability to communicate represents potential risks. In hunter-gatherer cultures of prehistory (and even now) the biggest threat to human groups in the foodchain was OTHER hunter-gatherer groups. Groups with whom some linguistic-level communication was at least potentially possible. Possible, but not always safe. One hunter stumbles across another over his cookfire: Richman the Stallard Hey, shove off! Lemme eat in peace. Don't you know who you're dealing with? I'm Richman the Stallard, of the Mighty Ung-- Wary of Lall: What? Who are you? What are you doing on our turf? Don't you know we don't like you people? Especially when you chew off your toenails and spit them into that awful porridge you make from stuff foraged from our collected garbage!? Stallard: I can't understand you! Can't you speak Psile? Listening to you is like trying to read smoke signals when someone's barfed into the fire! But I bet I could reverse-engineer your primitive parsing algorithm overnight! (Stallard disarms Wary, and turns him on his campfire spit, but, try as tortured Wary might to reveal his tribe's all-important parsing algorithm, Stallard learns no secrets worth having, and Wary's tribe remains safe. Stallard tries eating some of Wary's charred flesh as a consolation prize, but vomits into the fire at the first bite. People from Stallard's own Ung tribe follow the smoke to its source, assuming it must be those pesky Lalls again. They fall upon and kill Stallard before they recognize him. War is averted in the attendant confusion, remorse, and mourning.) Ahem. As I was saying... If you are born to a woman from one speech community, and grow up in another, your mother's language might be your "mother tongue," However, most of your socialization will happen among your peers and from the adults of your adoptive community. Your ultimate loyalty might lie with the people you grew up with, over your mother's original community. And your mother might never gain full acceptance in your adoptive community unless out of political motivation across BOTH communities. Without such political acceptance, it's safer to assume that those other communities are at best indifferent to the survival of your own, and maybe even hostile. In crunch periods such as ice-age onsets, droughts, famine, epidemics, and so on, differences in language and even dialect are clues to your "selfish genes" about who to save, who to let go, and even who to kill. (This is broad-brush, admittedly. In any community, there will be some tendencies to compassion for strangers, and even streaks of xenophilia rather than xenophobia. The default setting, however, might be a naturally-evolved guardedness about strangers who can't understand your language/dialect or speak it as well as you can.) This relationship of language to the potential for hostility is encapsulated in a linguist witticism. Having puzzled over the question of the difference between a dialect and a language, linguists finally arrived at this: A language is a dialect with an army and a navy. This isn't all Nature Red in Tooth and Claw. Obviously, just as evolution has provided for a language to be learned easily from a community, it has also allowed for learning more than one. Imagine if your mother taught you English from infancy, and because of that pre-occupation of your language centers, you couldn't learn Finnish--and you were starting kindergarten in Helsinki. Why bother? Given what must have been the huge prehistoric diversity of languages, and the continuous roaming of hunter-gather cultures, the POTENTIAL to be multilingual would generally have both individual and group-survival benefit over any hardwired-monolingual human strain. Inherently monolingual humans could at best only be able to fight or ignore other groups that they might also cooperate and marry into. Inherently monolingual humans would have individual- and species-level survival disadvantages. But the full benefit of the multilingual potential only goes to those who, by virtue of parentage from one community and upbringing in another, can be understood and trusted to a great degree in both communities. Adults who could learn foreign languages without effort couldn't be trusted by anyone, not fully. They could too easily run off and find protection in some foreign community. There might not be enough "linguistic surface tension" to hold your band together. Thus, in a very backhanded way, your adult difficulties with learning a foreign language are actually part of a group-level defense mechanism. (Don't you feel ever- so-grateful?) WITH LANGUAGE INSTINCTS COME INFORMATION INSTINCTS Now: what might the above hypothesis, if true, suggest about any possible human instincts about information? Especially as they relate to community membership? There must be more, here. After all, we don't instantly trust someone just because they speak our native language as well as we do. In some cultures that have experienced a high degree of internal dishonesty and hostility and a high degree of exposure to more beneficent foreign cultures, the reaction might be the opposite. This might be unusual, but it does exist. I would suggest the following (which is forehead-smackingly obvious): the closer we are to someone, the more willing we are to talk with them. The more we regard someone as a stranger, the more guarded we are. This too has its exceptions, but they are, again, rare. CHARGING FOR INFORMATION What does it mean for someone to charge someone else for information? This situation undoubtly predates money, but it's very doubtful that it predates language--if anything, language must have been quite highly developed before such a transaction could even be negotiated. The concept of "information about information" is crucial. Again, what does it mean for someone to charge someone else for information? This question is too simple to make sense. It has to be placed in a socioeconomic context before information instincts could possibly apply. Some missing social dimensions are: in-group vs. out-group membership, and whether the information is conveyed orally or by artifact, and whether the information is about in-group members or not, whether it's about either the speaker's or the listener's outgroup, or about an outgroup relative to both participants; and where the intelligence came from originally. Looking first at exchanges for which people don't charge: Two people who are in the same in-group, sharing information orally--the only inhibitions might be related to how the imparting of knowledge might affect subgroups. Gossip is an excellent example. If there are information instincts about gossip, they are, if anything, positively reinforcing. Both the speaker and the listener are getting a thrill, after all. But one might be more careful about gossiping about one's own family members to some in-group member outside your family. And one might be careful gossiping even within the family. The usual in-group/out-group picture is complicated by the subgrouping and dominance relations within and between families. Two people from different in-groups, orally: the speaker will consider whether the information is about in-group members, or could be used against an in-group member. Most people are considerably more guarded about this kind of information transmission. Two people in same in-group, via artifact. Concerns open up about possible future out-group exposure. The artifact might survive with some evidence of having been a medium of communication. And so on. The combinations are denumerable. The degree to which instinct and reason are engaged will vary radically depending on situation and recent events. I propose only the following rules: If a piece of information is valuable to all members of an in-group, charging in-group members for it will be seen as casting doubt on the vendor's in-group membership. If a piece of information might be valuable to out-group members, and holds any potential at all for use against the in-group, giving it away casts doubt on the donor's in-group membership. I suspect that these rules are not axioms, but can rather be derived from my hypothesized information instincts. I suspect (but can't prove, yet) that genetic-algorithm simulation models will bear out these hypotheses. Humans that didn't follow these sorts of rules had less survival potential than ones that did, and now they aren't around anymore in significant numbers. How does all this relate to software? Here is my theory. Because software, like speech, can be repeated with almost zero duplication cost, people have roughly the same instincts about software that they do about things they've been told. In the case of shrinkwrapped software, people tend to give it away relatively freely to their friends when they feel safe in doing so. Easy come, easy go. And it's just information--it's not like you gave away someone else's THING, right? Software publishers have fought hard against this mentality, but with only limited success. Basically, somehow, people don't LIKE being told they can't propogate information freely if it came from an outgroup and seems clearly useful and not particularly dangerous. At the same time, however, people fear viruses in software just as people fear contamination of the in-group values by out-group messages. They don't even have to be certain of any danger to act on the fear. USING LANGUAGE TO TRANSCEND SPEECH COMMUNITIES: LAW Maybe now it's time to start talking about Law, which is probably about as old as full-fledged language, but which in its formal properties resembles programming languages to a great degree. This is relevant to the discussion because, much as people fight over GPL and other licensing issues, nobody has yet to go to court over them. In modern times, law, and government, remain the ultimate outgroup. Yet they are also the ultimate guarantors not only of the value of money, but also of copyrights, of patent rights, freedom of speech and information (or the lack thereof) and finally, GPL itself. Even before any of those intellectual-property institutions existed (and yes, I believe money is an IP institution, the government being regulated-monopoly owner), there were laws about information and who you could pass it to, and when. (to be continued)