Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!zaphod.mps.ohio-state.edu!ncar!noao!arizona!gudeman From: gudeman@cs.arizona.edu (David Gudeman) Newsgroups: comp.lang.misc Subject: Re: The Universal Language (Was Re: Efficient Fortran) Message-ID: <24011@megaron.cs.arizona.edu> Date: 8 Aug 90 04:53:20 GMT Organization: U of Arizona CS Dept, Tucson Lines: 63 In article mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer) writes: > > Given this view of what a universal language is, the question becomes > ``to what extent can this hodgepodge be unified and streamlined?''. > The question becomes a quantitative rather than a qualitative one. >... >The problem isn't streamlining the various specialized notations >together, it's providing all the facilities for manipulating those >objects that are going to be needed. The point to unifying and streamlining (both terms chosen deliberatly for vagueness) is to reduce the cost of learning the language. Presumably a universal language is going to include a lot of stuff. It would be a good idea to minimize the size of the language by (for example) avoiding redundancy in notations. > ...finding > relationships among the various specialized features, subsuming them > with more general features. > >Yes, but if that "more general" feature doesn't map simply to the >specific features needed, it's not clear you've made an improvement >over including them both. If the more general feature doesn't map simply to the specific features needed, then it is not more general. >Such a language is doomed to failure. The problem can be expressed as: >Do you have hooks for dealing with sigournism, a field of study whose >subject matter will be discovered in 2012, and spring into existence >thereafter?... >English solves this problem by being "infinitely extensible"; Sorry. I thought it was obvious that any language that makes a claim to being universal will have to have some ability to evolve. Not just to add new features, but also to loose old, unused features. In fact, we may well have to loose the preconception that a language must have a single, unambiguous definition. Such definitions do not exist for natural languages and it may finally be shown not hold in the more formal realms either. Since Greek times there has been a prevalent idea among Westerners that given the knowledged and tools, the universe can be described with perfect accuracy and precision. This attitude came into its modern form with Newton's laws. However, there are areas where it is frankly not believable that this will ever be possible. Such areas include the behavior of individual molecules in a gas, and the behavior of individuals in a society. Modern physicists have even come to the conclusion that even in principle it is not possible to describe _anything_ with perfect precision and accuracy (both at the same time). Many linguists have given up the idea of rigid rules to describe natural language, and instead say that the meaning of an utterance is infered from general rules and the surrounding context. It might be that one of the discoveries that would lead to a "universal language" is a way to make a clear and useful trade-off between precision and accuracy. Don't ask me how this might apply to programming languages because I don't know... -- David Gudeman Department of Computer Science The University of Arizona gudeman@cs.arizona.edu Tucson, AZ 85721 noao!arizona!gudeman