Wednesday, February 19, 2014

My Proto Semantics is nearing a final form

Took a look at Wikipedia's definition of "semantics" today. It is a huge mish-mash of form and content, syntax, grammar, vocabulary, semiotics, linguistics, and language differences - all supposedly in pursuit of the subject of "meaning".
Very little of it seemed to have to do with what I consider narratives, or stories. Sure a Chinese speaker may have some different narratives than me, but we also must have many that are the same and are about the world around us and the insides of our minds. It is precisely those worlds that we share and the common narratives about them that ought to be the proper subject of semantics.
So I slice it and dice it differently from Wikipedia. Form is represented by a proto semantics, as per below. Content is represented by word meanings and the larger mystery: how do words get their meanings?

A PROTO SEMANTICS
Nouns
There are three kinds
 - person (me, or things I lend me-ness)
 - thing
 - location 

These are denoted by single letters or groups of things in parentheses. X, Y, etc.

Adjectives
Two kinds
-        feeling (attributed only to persons)
-        attribute  

These are denoted by A,B, etc. To express that a noun X has or feels an adjective A we write:
X__/A

Verbs
These involve a pair of nouns called actor and target:
actor\target
person
thing
location
person
love
understand
want
assign_value
see

go
indicate
find
thing
cause_to
act_on
compare_to
in
at
location
affects
contains
on
connect_to

To express that a noun X acts on a noun Y we write:
X-->Y
To express the idea that the same verb occurs in more than one part of a narrative, superscript the arrow like this ‘-->a‘.

Note that noun and adjective types are automatically converted by usage. To say “the dog loves his owner” or “Niagara Falls loves to see tourists at all seasons” lends personhood to these non-person nouns. Similarly we will be able to put attribute words in the locations of nouns (e.g. “red is shirt”). Although almost nonsense, such constructs do carry slight meaning.


Narrative Fragments, Connectors, and Grouping
Narrative fragments are:
·       noun
·       noun_/adjective
·       noun-->noun
·       two narrative fragments joined by a comma ‘,’. This is a connector that means ‘consecutive’.
·       two narrative fragments joined by a ‘::’. This is a connector that means ‘becomes’.
·       Any narrative fragment in parentheses. This means ‘treated as a noun’ or ‘treated as an adjective’ depending on the usage.
·       Any narrative fragment in square brackets. This means ‘implicit’ noun or adjective depending on usage.

Rule of precedence
For simple expressions: ’__/’   ‘-->’  '::'  ‘,’ .   
Otherwise use parentheses to avoid ambiguity. 

Update: Most of the arguments in semantics seems to have to do with whether words can be used in narrative roles that do not match the words' natural definitions. Duh! I realize that there is such a cacophony of nonsense out there about this subject, good new ideas will never be heard, unless they become the basis of commericially successful applications. This seems like a reasonable test, if evaluated in the long term.

Tuesday, February 18, 2014

Snow day for many people on the East Coast

I got to work ~6:30 and left shortly after lunch at ~12:45. I had a very hard time getting home only a few miles away. Now it has been snowing hard for 3 hours. I am happy to report that, over on my other blog, it looks like some people spent long chunks of time today reading about rock piles. Mostly about the Mayans in Georgia (good lord look at the photos from Track Rock Gap!) and arrowheads. But it is nice to offer value on a snowy day. Of course no one comes over here.

The geometry of mind

I end my paper on Best Models (while discussing reading as a sequences of frames of reference) with the question: "what is the data of this mental landscape and where are these frames attached?
The answer is in the direction of words and context networks. The geometric objects are narratives instantiated with such words. [Later on I start calling this narrative shape.]

Thursday, February 6, 2014

Google's sale of Motorolo Phones is well timed

It seems weird Google would sell the most American of brands, Motorola, to Lenovo. But for crying out loud: suddenly a couple days later Lenovo discovers a market for these phones in India? WTF?

Tuesday, February 4, 2014

The Star Spangled Banner

Hey, play this one when they get a 51st state (and use my flag):

Saturday, February 1, 2014

The idea of a machine being more intelligent than a human is meaningless

You know how these IQ measurements are nonsense cuz they pretend intelligence is a one-dimensional quantity? It is just as bad failing to understand that intelligence is a way of achieving human goals. A machine that was super good at achieving a non-human goal would just be another piece of mining equipment. No one would call that intelligence.