Toward An Alien Dictionary

This is a work in progress which I’m hoping a few people might find interesting and that interest might help further the concept.

Thais describe crows by the way they name them as Nok Ka where ‘Ka’ sounds like the ‘caw’ of a crow. They do the same for doves which are Nok Kow where ‘kow’ has a sound similar to the bird’s ‘coo’. Sadly not all the birds are categorized in the same manner but I think they are all predefined by ‘Nok’.

So would this be a good way to organize a naming scheme – by actually defining the basic characteristic of the nouns in a similar way to the ‘Animal, Vegetable or Mineral’ guessing game?
I believe it would be, but the first difficulty of adopting this method arises with choosing the categories… I think something a little more sophisticated than ‘Animal, Vegetable or Mineral’ would be required for a practical model.

Before we rush off into exactly how to define nouns it is probably worth considering all the rest of the words which a typical dictionary would require.

Now, I’ve spent quite a lot of time over the last couple of days chatting with a Lexicographer who tells me that he could probably manage to simplify almost any text using a core vocabulary of less than 3,000 words.

Although different languages use different syntax he tells me that all the major languages would probably fit into 3 sub divisions with the ‘Romantic’ languages being the most widespread.

As I see this wouldn’t make any real difference to our alien dictionary it as it is mostly about the method of sentence construction.

My friend seemed to be more interested in verbs than nouns but I argue that nouns are the building blocks that the other words either embroider or modify - but I agree that in some ways it is an arbitrary choice.

For me, the most alien thing I can think of is a computer because it has no ‘life’ as most people would define ‘life’ and no ‘senses’ as most people would define ‘senses’ – although it can be argued that they can ‘see’ and ‘hear’ and, to a limited sense ‘feel’ as, say, by detecting the difference in pressure or temperature on a suitable sensor.

What I’m trying to get at is that the easiest way to define something concrete is to point to an example of it than can be examined by the senses - but this requires not only senses but a collection of basic differences that such ‘senses’ can sense and some method of defining the scale and magnitude of the properties that these senses can detect.

With a computer you could not really offer it any sense of ‘good’ or ‘bad’ that would have any intrinsic meaning to it; i.e. you couldn’t use a certain sequence of words to inspire hope, or joy, or fear because a computer does not have such ‘emotions’ and, quite possibly an ‘alien’ wouldn’t either, or at least would probably have a completely alien ‘set’ of emotions.

The question is ‘would developing an Alien dictionary help reduce the coercive power of words or is it that humans have ‘emotional buttons’ that can be pressed regardless of the method of pressing them?

Anyway, let’s return to the design of our Alien Dictionary.
Obviously an alien who knew nothing of Earth or its languages would benefit from a language in which each word was actually a signpost to its meaning – in a way similar to Nok Kow which signifies a bird which makes a particular sound.

The problem here is that a sound relies on the sense of hearing and the written sound does not give a clear indication of the actual sound.
So, how should we deal with descriptions of a word if we don’t know which, if any, of our senses the alien will have?

Given a large enough data base you could actually devise names that were true signposts to pictures, sounds, and ever more detailed descriptions of the noun.
So, say dove would be Nok which points to Animal (bird) and Kow would be an additional signpost to the specific bird from which any added describing words could point to an even more specific description of such a bird.

So, would a kind of hypertext, multimedia solution be a realistic approach?
Well, if the dictionary database is on the Internet I can’t see why not.

The computer being used would have a very limited subset of the database but would reach out to the online database for added detail if, for instance, bird wasn’t a sufficient description and, say, the sound the bird makes or the fact that doves can be associated with peace was important.

Mykle has already suggested a basic method of building concepts by taking nouns which can ‘own’ adjectives, verbs etc so you get say, car - which then can ‘own’ a size, a value, a colour, capability etc. as descriptive properties (big, expensive, red, fast)…
http://www.abctales.com/story/mykle/introduction-classylang

If such a system were to be adopted then perhaps Nok would not be a basic enough description and it would need to be noun, animal, bird, Nok in that each word would have to start with a basic description of its hierarchy, at least in its first appearance in the text.

Mykle adds extra word types to facilitate non human word types used in computer related languages which would greatly aid the language when talking to machines.

I've heard that all creatures will soon be categorised by using their DNA ...
Call it a DNA digital Dewey Decimal System for all life on Earth.
http://uk.reuters.com/article/2010/10/29/us-dna-idUKTRE69S57G20101029
It will be very interesting to see, in detail, how this is implemented and whether or not the principle can somehow be extended...

I've been thinking about the "DNA digital Dewey Decimal System" and although it would be fine for identifying biological life by taking a DNA sample it is hardly a great way to classify species in a dictionary.

I think a much simpler classification based on obvious characteristics would be of a lot more practical in day to day usage.
Could we implement a simple system of classification?
Does anything similar to a binary choice system of classification already exist?

Well, yes, in a way it does.
The BBC has a KS2 page which offers :-
“Plants are divided into two groups, flowering plants and non-flowering plants.

Animals are divided into two main groups. Animals that have a backbone are called vertebrates. Animals that don't have a backbone are called invertebrates.”

A good start, however, it then goes on to divide Vertebrates into fish, amphibians, reptiles, birds and mammals which would be difficult to express in a binary manner.

If, for instance, you tried to divide them into say ‘air breathing’ or 'not' then fish would definitely be not, but what about Amphibians which are vertebrates that spend part of their lives under water (breathing with gills) and the remainder on land (breathing with lungs)?

I suppose you could define a system of binary questions which could get around the problem because then an Amphibian would have a yes for breathing water and a yes for breathing air.

However, if this system were to be adopted how many such questions would it take to define all known species?
If it can be done in less than 60 questions then I think it might be a viable method of definition.
The trick would be in finding the right questions.

---------------------------------------

Maddan very kindly gives a very useful link (see his post below) which gives some examples of earlier attempts at using a similar principle to the one I have outlined earlier in this piece.

Since I have never heard of any of them I presume that none of them caught on. So, why do I think another attempt might?
Well for two main reasons; computers and the use of translators to either ‘complile’ or ‘interpret’ from a local language to the ‘Alien’ one.

The Lexicographer I was chatting to pointed out that a good dictionary includes examples of the use of words and this can be difficult because different dialects can have different fads and fashions.

If you consider the expression used above of ‘caught on’ it isn’t exactly what you might guess if you were a stranger to the English language.
Similarly, do you ‘take’ a bath or do you ‘have’ a bath?
If you do take a bath where do you take it to?
Is having a bath like having a woman?

He also told me that words like ‘look’ are difficult word to define concisely and I can see that they probably are…
So how do we get around such problems?

The choice of translating the local language to the Alien one allows much of the difficulty of usage to be dealt with and has the added bonus of allowing the actual Alien words to be transparent to the user.

As Mykle points out a lot of effort would be needed to get the Translators to work well and, I suppose, they may never be flawless - but then what is?

I expect that a lot of the early attempts might turn out fairly interesting poetry as quite often poets use unusual combinations of words to force the reader to think and it is inevitable that such combinations would be fairly common in early translators.

The fact that computers are fairly ubiquitous and very powerful nowadays offers the ability to use a fairly sophisticated method of creating words and linking them together…
so no need for an alphabet as we would think of one but rather a set of defined categories modified by position and value probably using a 64 bit word expressed in Octal or Hex.

---------------------------------------

I'll get back to this later but in the meantime any thoughts would be really welcome.