[texhax] help with identifying some macros

Lars Madsen daleif at imf.au.dk
Sun Oct 18 13:24:24 CEST 2009


Toby Cubitt wrote:
> On Sat, Oct 17, 2009 at 11:08:48AM +0200, Lars Madsen wrote:
>> Using standard LaTeX macro names are not that good as they often
>> hide the true meaning of the equation.
>>
>> For example in topology \cong often means isomorphy, so
>>
>> X \isomorph Y gives much better meaning than X \cong Y
> 
> In this instance, sighted people reading the final typeset version are no
> better off when it comes to interpreting the maths: they see the visual
> output of "\cong" and have to decide from context whether that symbol
> means isomorphism, some form of approximate equality, or whatever else it
> gets used for.
> 
> So if authors consistently defined a separate alias (such as \isomorph,
> \approxeq) for each distinct use of the same symbol (\cong), interpreting
> the maths would become easier for everyone with access to the LaTeX
> source, not just blind people. I seem to recall the amsmath guide
> recommends doing just this.
> 
> I guess the downside for blind people would be that instead of having to
> learn a (relatively) small set of standard symbol names, there would be a
> large set of command names to interpret, many of which would produce the
> same visual output. 

not really, because the naming should show the mathematical meaning of 
the symbol, not describing how it looks. Such that when reading the code 
the equation starts to resemble how it is actually read out loud.

Think of the scrolling text green from the Matrix movie.

> It's probably impossible to define standard commands
> for every possible distinct use of the same mathematical symbol (research
> papers for example very often have to define new notation as needed for
> new concepts). So some command names will end up being author-defined,
> and there's no guarantee everyone will consistently choose the same name
> for the same thing. In the worst case, interpreting a command would
> involve finding the macro definition, figuring out what standard command
> it produces, and *then* figuring out it's mathematical meaning from the
> standard command anyway.
> 
> I wonder if we're not better off as we are, with (more or less) a unique
> command for each unique symbol.
> 

The problem is that the meaning of the expression gets lost, we are 
doing a visual representaion of an equation, but the underlying meaning 
is lost.

For example where I work some people use the notation

X \sim N(0,1)

to say taht X is a stochastic variable which has a normal distribution, 
whereas

x \sim\sim N(0,1)

is an observation from such a variable.

Reading that makes no sense at all, but using better names for symbols 
it becomes much more readable.

X \DistAs N(0,1)

x \ObsFromDist N(0,1)

One should also think not jsut of the reader, but also co-authors, using 
better more descriptive names for macros makes cooperation a lot easier, 
and perhaps we will get rid of a lot of the mess people are doing be 
using there very short non-descriptive macro names, that I know scare 
most journal editors.

/daleif






More information about the texhax mailing list