Re: Babbage/Menabrea: Analytical Engine

From: Marinho Francis-Oladipo (fom198@ecs.soton.ac.uk)
Date: Mon Feb 19 2001 - 11:27:58 GMT


<http://psychclassics.yorku.ca/Lovelace/intro.htm#
<http://psychclassics.yorku.ca/Lovelace/menabrea.htm#

MARINHO:
This is an afterthought to my earlier comment on menabrea's conclusion on
the subject of intelligence, suggesting that even though it was not
necessarily intelligent, the Analytic Engine executes conceptions of
intelligence. Earlier on in the paper, he divides the mathematical sciences
into mechanical and reasoning sections. The reasoning section according to
menabrea is the region of

> MENABREA:
> pure intellect that which depends on the reasoning faculties.

MARINHO:
If this is accepted, intelligence as I interpret it is mathematical
calculation with an aspect of reasoning. In other words computation
with understanding. I gleaned from the paper that Babbage wanted his
machine not to display intelligence, but instead to be able to perform
calculations of any type. It appears to me that from the details of the
machine as described here by Menabrea, some people would have
eventually stumbled onto the arguement about whether intelligence was
at play here or not. I therefore think that Menabrea indirectly tried
to dispel such notions by emphasising that the machine only displays
conceptions of intelligence, and is more a tool dedicated to performing
mindless calculations using specified algorithms, thereby leaving human
beings more time to do the intelligent reasoning.

With respect to the fact that we are not informed in the paper how
division is implemented, I feel it is quite an important issue. A
machine is being described that can perform absolutely any mathematical
calculation. I find it interesting that there is no knowledge or
insight at that stage as to how the implementation of one of the basic
mathematical operations would be handled. It can obviously be done
using addition/subtraction but the author wants the reader to assume it
can be done directly. Surely, if implementing one of the foundation
operations cannot be described, there is no need proceeding with the
rest of the machine especially considering that every function is
perceived as being able to be reduced to a combination of the four
basic ones.

Finally, it is important to remember that all this was an idea and had
not actually been implemented even though there were plans and drawings
all over the place. Most of these claims were made while the machine
was still in the idea phase and so could not be backed up with evidence
which must have been very frustrating for Charles Babbage.
Nevertheless, he made claims that should and could not have been made
without having tested a working system. One such claim as I mentioned
earlier was the one pertaining to the speed of the machine by being
able to perform a multiplication in three minutes. It is interesting to
note that such a claim was made and yet 31 years after the lecture in
Turin, Italy, when Babbage died, a working Analytic engine still had
not been built and even after his death when his son carried on his
legacy, the mission was still unsuccessful. I'm however not trying to
imply that Charles Babbage was a "Bad Ideas" man and his machine was
all talk, instead i'm just laying open ground for thought on whether
such a machine would have eventually been implemented. A lot of the
hold up in implementation was due to the lack of trust and hence
funding for the project but according to the Lovelace/Menabrea intro, a
working Difference Engine was eventually built in 1854 and it is clear
that Babbage had some very sound ideas. So i believe that given the
right backing and support, Babbage's dream might have been realised
early enough for him to see it and be proud.



This archive was generated by hypermail 2.1.4 : Tue Sep 24 2002 - 18:37:15 BST