[UEB Maths] minutes from our teleconference

[Janet Reynolds]

Here are the minutes from the teleconference. The CD worked really well,
I was able to play it on the PC and precis each comment. I've put each
of our summary comments in full at the end. A useful exercise reading it
through, you can see us trying to balance all the factors.

I found the braille version of the economics book we brailled here in NZ
but no longer have the print. It didn't have functions as such, just
word fragments used as variables. For example ENP sub 2. When doing this
in Nemeth we decided to use double caps to make it clear that it was not
three variables multiplied together. Technically it is an abbreviation
being used as a variable, I can't remember whether it was italicised in
print or not.
Janet
 =20
Teleconference: 14 July 2005 9 am (NZ time)
Present: Janet Reynolds, Joe Sullivan, Stephen Phippen, Bruce Maguire,
Helen McMillan
=20
JR: introductions
JR: read through example 1a, pausing on 7sin 30
JR: the current draft of section 9 says the space before the argument
can sometimes be omitted but I'm feeling now that that if we don't have
function indicators that space before the 30 is definitely needed.
SP: if we don't have the function indicators, then we need the space
between the 7 and sin as well between sin and 30
BM: Agree they should be symmetrical, no spaces or both spaced
HM: like the space after but don't see need for one before
JR: Helen and I both used to seeing it that way in Nemeth
SP: Can be ambiguous if the 7 was an algebraic letter
JR: Ambiguities like that not common
BM: Might be on the increase, particularly in economics
BM: Of the three examples, like 1a least
JR: read through example 1b, pausing again on 7sin 30
JR: written here with the ch separating the 7 and the sin, and the dot 3
separating the sin and the 30
SP: according to the rules I suggested the dot 3 wouldn't be needed
here, the ch could be terminated by anything other than a roman letter
or a fullstop
BM: I wonder how easy that would be for students to implement when
writing, in practice simpler to have rule that you always use the dot 3
SP: dot 3 just acts as a mental separator, perhaps not needed when
something like a number sign or greek letter indicator serves the same
purpose
JR: in trig the angle is often either a greek letter or a capital
letter, would you need the dot 3 before a capital?
SP: capitals are allowed in the function name, eg Log, but you could say
initial capitals don't terminate the ch
SP: Try to keep it compact as well as reflecting mental processes.=20
JR: When I first drafted the function rule I used dots 56 as a separator
in a similar way but Antonnette felt the 56 was inappropriate used that
way.
SP: The rule I drafted is similar to the current BAUK rule, I don't want
that to be the reason for it to be pushed, but it's perhaps why I'm more
comfortable with it
BM: Still worried about a rule that might not get followed in practice,
particularly if needed for back translation. Of the three alternatives
prefer 1c, because there the dot 3 is more natural.
SP: While we are still looking at 1b, you could argue that opp and hyp
should also have the ch indicators, because they are not variables in
italics
JR: One of my worries is that the rules for when the ch indicator should
be used need to be very clear, not all transcribers understand what a
function is. If it applied to all sequences of letters not in italic
then what about abbreviations? Danger of having complicated rules for
the transcriber and too much noise for the reader that comes between
them and the teacher in the classroom.=20
SP:  don't you think the braille reader is disadvantaged reading opp and
hyp without the extra visual cue=20
the sighted reader is getting that these are printed differently
BM:  How would BAUK write opp over hyp?
SP:  the ed sign in BAUK is a general mechanism for recognizing word
fragments, though you wouldn't use it for apples plus oranges, BAUK
rules rely on context which is harder in UEB. In UEB the simplest way to
distinguish fonts would be to use italics for all the variables but that
would be too much. This ch indicator gives us a way to indicate the font
change just in those situations where it is mathematically significant,
which may be wider than functions, we need a general rule that is clear
for transcribers.
BM: In some teachers notes you may not get the variables in italics
JS:  Also when functions are written in say computer programs like
Fortran
SP:  So in those cases best not to use the ch indicator
JR: I disagree, would prefer to see at least the common functions listed
in the transcriber manual always brailled the same way, regardless of
how written in the print. In a maths setting the braille needs to show
the structure of the maths, not the inconsistencies of the print,
similar to our advice on spacing. Helen, do you agree?
HM: Yes I do agree but am having trouble with the phone connection which
keeps fading in and out.
JR: what are your current feelings about the function indicators?
HM: Understand the arguments for them, like the idea of them indicating
a font change, might help transcribers who find it hard to recognize
abbreviations as well as functions.
BM:  So we think of them as some sort of font indicator rather then a
function indicator
SP:  Could go in Section 1 where we mention not using italic for
variable
BM: If we think of it like that then presumably we couldn't have
alternative 3, with s for sin. (example 1c)
SP:  could still be defined as "contracted" forms of say sin, cos, tan
and log.
BM: but you might have a function called s
SP:  Unlikely to have a single letter function, but that would be a
reason not to have the contractions
BM:  I liked alternative 3
SP:  If we did define the ch indicator by font we might regret it,
because you do get other letters like matrix names, which might not be
in italic in print
JR:  where do we draw the line - if we had something simple like 6
metres, do we really want to put indicators round that metres?
SP:  That's why I'm thinking of word fragments, which would be more than
one letter so wouldn't include single letters
JR: Feedback from teachers here is a plea for consistency, if we always
show sin as ch s,  would you then lose a knowledge of what  s, c and t
stand for?
BM: You would soon get used to what ch s stands for, even if you weren't
as aware of how it looked in print, true for other things in maths, like
the fraction layout. Backtranslation could still give=20
JR:  Thinking of the struggling student who has trouble with the concept
of trig, is it easier for them if the braille says sin when the teacher
says sin?
BM: Interesting to see if that has ever caused a problem for Australian
students using ed s. Could ask Josie Howse. It's hard to know if people
just get used to whatever they're given.
SP:  If people who are actually using it say it's not a problem than
that perhaps counts more than people who are not using it saying it
could be a problem.
JR: Fair comment
JR:  Other feedback is that the rules need to be very quite tight and
very simple,  we need to decide if this indicator applies just to
functions like sin and cos, or whether indicators would be used for two
or more letters that are some sort of mathematical entity, we'd have to
draft the rules quite carefully.
BM:  Also need the flexibility to cater to the needs of individuals,
like the way we say that spacing round operations signs is a useful tool
for young students but can be dispensed with when they progress.
JR: I used to think that, but feedback here says it doesn't matter so
much what you decide as long as it is black and white, because people
doing the transcribing are not always knowledgable enough to apply the
flexibility sensibly.
BM: That's an ongoing training problem for transcribers
HM: the problem here is that we often split a book up between several
transcribers, so don't want different interpretations within the same
book.
BM: those are process issues for a production house, I've learnt that
there has to some kind of editorial input
HM: but the more flexibility, the more room for error
JR: Perhaps the guidelines can give a strong recommendation but make is
clear there is also room for small amount of flexibility.
SP: the guidelines in Section 1 on spacing follow that sort of approach,
it gives advice on spacing but makes it clear that the code design works
regardless, which is one of the design features of UEB.
BM:  similar with email addresses, code works either in contracted or
uncontracted, but that's where I would make the guidelines more
prescriptive, so everyone does it the same.
JR: time passing, let's look at the other examples
BM: How long do we have
JR: Aiming at one hour so 20 minutes left, but could go over by 10
minutes or so
JR: read through example 2b, pausing on cosec squared.
JR: where should the dot 3 go, the function being applied is cosec
squared which is why I put the dot 3 before the theta
SP: you might not need a dot 3 if you consider the ch  being terminated
by the superscript sign but if you did have a dot 3 I would put it after
the last c of cosec
JS: I would see it that way as well
JR: so would I now, after our discussions on word fragments
JR: I'm slightly worried about ambiguities because s superscript could
also read as s followed by the in contraction, so sin squared x, if you
use the s abbreviation and no dot 3, looks rather like sin 2x.
SP: but the superscript would terminate the function
JR: but there's potential ambiguity for a reader who is not thinking
about those rules
SP: they'd be in the mindset of ch s being one object
BM: I agree as far as the reading goes, more bothered by the writing,
and how well it turns back into print
SP: If people put too many dot 3's in when writing would it still
backtranslate alright?
JS:  It would probably backtranslate okay but it's whether the maths
student needs to be thinking more about the maths, and less about these
transcribing rules. With the backtranslating, the apostrophe could be a
problem
SP: If there was an apostrophe you could use two dot 3's, one to
terminate the ch.
JR: You could imagine functions with apostrophes, but they'd probably be
primes, so you could use the lower g prime sign.
BM:  Still feel it may be simpler to always include the dot 3, as
teachers will probably tell students it's safer to include it anyway
SP:  To me it just seems clumsy to include it in 90 percent of cases
where the number sign or greek indicator psychologically gives you that
break.=20
JR: But then it's tempting not to use it before a capital letter like
sin A, and that would be wrong
SP: Could modify the rule so initial capitals are allowed
JS: Can't think of a real case where that would be a problem but
thinking of the future there might be one
HM: So rule would say you only use the dot 3 when the function is
followed by a lower case letter
JR: Which I guess is how I have section 9 written at the moment, but
with a dot 3 instead of the space. I must say I quite like the idea of
using the dot 3 all the time because it's at the left of the cell and
gives half a space before the argument
HM: but not in the case of that cosec example where you have a
superscript
JR: Yes in that case the dot 3 is earlier and the 2 of the squared is up
against the theta, but I suppose a reader would be used to how these
superscripts are written in UEB and wouldn't read that as 2 theta, they
know it's squared.
BM: I'd be interested to take the rule as Stephen is suggesting it where
there are cases where you need a dot 3 and cases where you don't and ask
some people here including teachers.
SP: Perhaps as a result of this session we could have a longer list of
examples that we could put up on this list, following the suggested
rule, so we can all look at them, it's the feel of the Braille that will
be important.
JR: I could take something like that to a meeting I am having with some
teachers and transcribers in August, and see what they think, though
they'll just say "why can't you spell them out".
BM: Spelling out takes up a lot more space which is a problem when you
are having to write things in class.
JR: Can I also check in this example whether have got the sin to the
minus one correct, do you need those Braille grouping signs?=20
JS:  This was revisited by Committee 2
JR: And it was felt that the minus signs occur on the own, like in
Chemistry.
SP: So you would need the grouping signs, that was the deal
JR: So in that one you'd have the ch indicator then s then the
superscript (if we didn't have a dot 3), and that couldn't read as sine
because the reader is used to ch s as an entity,  then the minus one
would have to be enclosed with the grouping signs.
JR: read through example 3c, pausing on absin C
JR Shows the advantage of the function indicator to separate the ab from
the sin. Also, I threw in the suggested grade 1 indicator on this
example, the ing sign. How are people feeling about that one now, I must
admit I don't like it.
HM: I don't either
BM: Having read more now I'm no longer opposed to the three 56's
SP: I would say "don't go with this" for different reasons. Because it
is so easy to use it might get used too much
HM: What do people think about the passage indicator on a line of its
own, perhaps without the dot locator. I rather like using it that way.
JR: I showed it to some teachers here and they quite liked it, one
student suggested also having it at the end of a previous line, which
also might be okay.
JR: Looking at the way I used the indicators in 2a and 2b, surrounding
the whole page with the passage indicator, would you have broken that up
more?
HM: Yes, I would have opened them after "Solve the equation" each time
JR: Would you have come out of grade one mode for "Now" and "So".=20
HM: No, not for small words like that, but maybe for each question in
case the teacher takes them out of order.
SP: the use of passage indicators over a large chunk of text and
equations does look peculiar, wouldn't want it written into the rules,
different countries may have preferences. Similarly with the bit in
Section 1 about equations being in cell 3 overrun in cell 5, better not
written in there - would use cell 5 and 7 in the UK
JR:  Perhaps we could change that to read: When an expression is set
apart, the braille format should indicate this in way by suitable
indentation, for example cells 3 with overruns in 5 or cell 5 with
overruns in 7.
JR: what about the continuation indicator
SP: The UEB sign is actually designed for computer programs where each
print line is an entity, whereas in Maths you are more interested in the
break of the braille line, not the print line
JR:  So perhaps we reword it to reflect that
SP: but wouldn't you need a different sign, as it's a different concept
JS:  There might not be any harm in using the dot 5 for this too, but I
agree it's not how we defined it
SP: the current wording implies that it is the print line break that
matters, but really it's the braille line that matters
JR: In NZ we probably wouldn't use it, but how about: Whenever a
mathematical expression will not fit on one Braille line, the
continuation indicator can be used to show that it continues on the next
line
BM:  More of a coding issue that can be left to individual countries=20
JR: Would it be better to leave it out altogether
SP: If we are redefining the use of dot 5 it does need mentioning
JR: I will leave it in for now then, but reword both that and the set
apart bit about the overruns. Are people comfortable about the advice on
using the indicators=20
JS: Yes, but have we discussed whether we would use the abbreviations, s
for sin etc
JR: In BAUK there is a longer list, and thing like cot are upside down,
I suppose for cot you could have ct.
SP:  Not if we had the single letter rule, maybe a restricted set like
sin, cos, tan and log, maybe e for exp.
JR: Exp isn't very common these days and you tend to see ln for natural
logs, though if you had say log base a you could use ch l subscript a
JR: Stephen have you time to put some examples together
SP: Yes, I could find some text
JR:  what about brackets, I guess you wouldn't need a dot 3 for say sin
(x)
SP: No because the bracket would terminate the ch
BM: Have you access to any economics material, interesting to see how
this system applies where you have lots of equations with words in.
SP: I'll try to find some examples thing, when you get on to these sort
of cases that makes me think twice about how good an idea it is, need to
come up with a general rule, which somehow encompasses all these sorts
of cases in some sort of general way, but include cases like apples and
oranges or where the indicators are better not there.
JR Or abbreviations, =20
SP: Cases where words are embedded among other mathematical things,
rather than just next to an operation sign. Thinking about how
computable all this is, we need to make sure we are not digging a hole
for ourselves in terms of being able to compute it. Go along with
Bruce's idea that although we want to give guidelines we don't want to
frame the rules flexibly enough so that if a translator does something
it's actually wrong. For example, if the translator just puts sin x
that's valid as far as UEB goes.
JR:  So if a teacher aide for example just puts some trig straight
through DBT, there won't be any ch indicators, we're creating something
where human intervention is necessary
SP: Unless of course the translator recognizes sin as a function, or
it's marked up in some way in the input text, or scanned by OCR that can
turn it into MathML
JR: to wind up could we have some summary comments. Stephen is going to
put together some examples following the  ch dot 3 rule as we've been
discussing it but where do people sit with it at the moment.=20
JR:  For me I'm still uneasy about this because it's human intervention
and it's a bit different from what's in print, but I can see all the
arguments for it and I can probably go along with the majority on this
one.
JS:  I have the same uneasiness but certainly we don't want anyone
writing sinx with no space or other indication that one is an argument
and one is a function so we've got to find a way not to have that
happen. If we base things on font we can regard this as a shorthand way
of indicating a particular kind of font or a font exception as it were
in the midst of mathematics, on that basis I'm okay with it. I do start
to worry a bit more when we talk about abbreviating them, sin to s,
which is like a contraction in a grade 1 situation which is a bit mind
bending given the normal trend in UEB to only contract in grade 2.
HM: I'm not fond of the contractions either, I'm not sure they save
enough space to make them worth the effort. I don't mind the ch and the
dot 3. The rule that covers capitals and lower case letters, that's fine
and I think it might work and if it's going to work with translating
tools that would be good too. As a transcriber it's one extra thing to
learn but that's not a problem. So I can go along with that. I would
still prefer to have it the way we did it now with the spacing just
between the function and the argument, I don't think we needed a space
at the front, just between the function and the argument.
BM:  I certainly think we need something  other than the current
official position represented by the first example. There are too many
spaces and the spaces are not symmetrical and it's too spaced out. And I
think it just adds to the perception that UEB takes up a lot more space
in math and will be more cumbersome to write. So I certainly think we
need a system more like the dots 16 and dot 3. I would prefer to
abbreviate sin cos tan but I suppose that's because historically that's
the way we do it in Australia. But I'm less wedded to that than to the
idea that we need something like the indicators.
SP:  I've put this forward as a technical solution. I still think it's
my favourite solution to this. I think there's a problem that needs to
be solved and this is the best solution I can think of. If you weren't
going to use these indicators I think you would need to space evenly
before and after, I don't like having things like a sin x written a s i
n space x, I think that's logically incorrect. If we weren't going to
have this rule I would have a rule where you've got to space everything,
and looking at it that way I think these indicators make it more compact
and also give you more power in terms of representing what is really
there in the print because those are genuine distinctions. So it's a
stronger and a neater solution to the problem. But I'm happy to be
dissuaded.
JR: Thanks for that. It looks like the most likely consensus would be to
use the indicators with Stephen's rules but without the abbreviations,
but none of us are really sure yet and we need to look at some more
examples. We need a deadline to get some examples up, is two weeks okay
Stephen.
SP: Hopefully yes, perhaps an old school book with trig in and some
economics with word fragments.
JR: I will look for some economics too. We'll try and go for consensus
on this one using email but if we're having difficulty we could always
go for another conference call.
BM: Could we consider using skype, I use this successfully.
JS: Better with broadband, will put url details on the list so people
can try.=20
JR: perhaps Bruce and Stephen could try communicating, Bruce with
broadband and Stephen with dial up.
JR: We've made lots of progress today, thankyou to everyone.
=20