[Fsf-friends] No patents for mathematical or business methods or computer programs per se or algorithms in India

Ramanraj K ramanraj.k@antispam.org
Tue Dec 6 18:34:07 IST 2005

Rebentisch <arebenti at web.de> wrote to <in-help at ffii.org>
pointing to the following article:

India: Patentability of Softwares in India
30 November 2005
Article by Manisha Singh Nair

,----[Manisha Singh Nair wrote:]
| The Patents Act refers to computer programs in Section 3, which deals
| with inventions that cannot be patented. According to S. 3(k), a
| computer program per se is not patentable. This makes us think about
| what the term ?per se? stands for in this context.
| According to the Webster?s Encyclopedic Unabridged Dictionary, the
| term ?per se? refers to "by, of, for, or in itself; intrinsically". If
| we are to use this definition we can well assume that the software as
| such cannot be patented. But don?t the same words of the provision
| tell us something more- that if the claimed invention is some thing
| more than?mere? software, it is patentable?

The Patents  Act declares  in Section 3,  among other things,  that "a
mathematical  or business  method  or  a computer  program  per se  or
algorithms;" are not inventions within the meaning of the Act.

Under  Section 7,  every application  for a  patent shall  be  for one
invention only.  It  is fairly obvious that an  invention that relates
to a mathematical  or business method or a  computer program in itself
or algorithms are not patentable in India.

We  only   need  to  investigate  the  situation   when  an  invention
incidentally includes a computer  program or even mathematical methods
as part of a single invention.

The  Copyright  Act defines  "computer  programme"  and "computer"  as

"computer programme"  means a set of instructions  expressed in words,
codes,  schemes or  in any  other form,  including a  machine readable
medium, capable of causing a  computer to perform a particular task or
achieve a particular result;

"computer"   includes  any   electronic  or   similar   device  having
information processing capabilities.

The "computer"  then refers  to devices having  information processing
capabilities, that may be

[1]  electronic: Without doubt,  laptop computers,  desktop computers,
mainframes and  other electronic devices  with "information processing
capabilities"  fall within  the  meaning of  "computer".  The  general
scheme of a generic computer could be as follows:

|       |               |        |
|Input>==> Information >==>Output|
|       |   Processing  |        |
|       +----^v----^v---+        |
|       |     Storage   |        |
Fig 1: An electronic computer

[2] The definition also  includes "similar devices" within the meaning
of "computer".   That brings within  the purview of the  definition of
"computer",   mechanical  devices   like   abacus,  slide-rules,   the
Analytical Engine of  Charles Babbage and any other  device that could
do information processing.

| o o o o o o o o o o o o o |
| | | | | | | | | | | | | | |
| | | | | | | | | | | | | | |
| o o o o o o o o o o o o o |
| o o o o o o o o o o o o o |
| o o o o o o o o o o o o o |
| o o o o o o o o o o o o o |
Fig 2: Abacus - a mechanical computer

Now, consider the following:

    A set of instructions to operate an abacus.
    A novel method to operate an abacus.
    A  set  of  instructions  to  operate an Analytical  Engine.
    Emulators  that  mimic  an  abacus  or Analytical  Engine.
    Programs to operate electronic computers.

In  all  the  above   cases,  the  instructions  are  instructions  in
themselves to  enable computation on  computers and there could  be no
confusion whatsoever that they are  "computer programs per se" and not
inventions withing the meaning of the Patents Act.

It is true that computers are universal machines, and we could express
many  inventions  as  computer programs  in  the  form  of a  list  of
instructions.  Many  manufacturing processes could be  expressed as an
abstract series of steps, that  at first sight, would appear as though
the  invention  could be  fully  implemented  as  a computer  program.
Closer  analysis  will reveal  that  "computer  programs  per se"  are
clearly distinguishable from  other inventions dealing with industrial

According to Patrick Henry Winston, "Artificial Intelligence" is study
of the  computations that make  it possible to  [1]perceive, [2]reason
and [3]act."  The dictionary meaning of "Intelligence" is "the ability
to gain and  apply knowledge and skills".  Only  those inventions that
exhibit  intelligence  are patentable  as  Section  3(a) excludes  "an
invention  which  is  frivolous  or which  claims  anything  obviously
contrary to well established natural laws".

Intelligence is a natural attribute found exhibited in life forms, and
specifically  required in an  invention, to  be patentable.   ("AI" is
plainly,  artificial "man  made"  intelligence, where  the measure  of
intelligence   is  only   incidental).    Intelligence  exhibited   by
inventions may be analysed as follows:

Intelligence/    |[1]Perception  |[2]Reasoning   |[3]Action
Invention        |(input)        |(logical steps)|(output)
Computer         |input devices  |information    |output devices
                 |               |processing     |
Abacus           |push beads with|instructions   |read/feel position
                 |fingers        |               |of beads
Bread Making     |flour,water,   |step by step   |bread
                 |salt,heat      |mnf process    |
Semifab Unit     |silicon,gold,  |VHDL & process |Microprocessor
                 |topography,etc |               |
Humans           |ears,nose,eyes,|brain,         |arms,legs,speech,
                 |tongue,skin    |nervous system |getital & excretory
Robots           |vision,hearing,|software       |move with wheels,
                 |IR,..          |               |speakers, monitor ..
Fig 3: Analysis of Inventions based on Intelligence exhibited

With  a computer,  that by  definition means  only devices  capable of
information processing,  it is neither possible to  input flour, water
or salt nor get bread as output.

The manufacture of bread may involve the following:
    Step 1: Take one measure of flour
    Step 2: Add two measures of water
    Step 3: Mix flour with water and add 1/100th measure of salt & yeast
    Step 4: Bake in oven for 25 minutes at 200 F

The  above  involves "mathematical  methods"  at  each  step, and  the
invention  could  never  be  described  without  use  of  mathematical
methods.   Under Section  10 of  the Patents  Act, the  patent  may be
supplemented with models  and samples which however are  not deemed as
part  of   the  specification.    Every  invention  could   use  other
inventions, mathematics, laws of physics, computer programs, and other
aids to  describe the  invention coherently, but  they would  never be
deemed  as  part  of  the   claim.   If  the  claim  relates  to  just
mathematical  methods or  computer programs  per se,  then  they would
simply be rejected  as not being inventions within  the meaning of the

It is very clear that  "Software patents" are illegal and unlawful not
authorised by law in India.

Literary  works upon  patents are  one of  the chief  benefits  of the
patent system.  For example, let us take Posilac 1 Step (US patent no:
4,985,404).  This is used to "increase" milk yield of dairy cows.  The
patent  expires on Jan,  15, 2008  and until  then, others  cannot use
Posilac  without  permission  from  the  patentee.   However,  nothing
prevents  physicians  or scientists  from  researching  the patent  to
publish articles or make other  decisions about the effects of Posilac
on cows and humans.   Research scholars could write simulator programs
to explain  the invention or analyse  its effects, without  in any way
infringing upon the patent.

A final  note: The  line between hardware  and software  is vanishing.
Using a  language like VHDL, it  is possible to  describe very complex
hardware as program constructs, and also implement a solution that may
defy  the  distinction  maintained  between "computer"  and  "computer
program".  We  may not be  very far away  from the day  when engineers
come with a machine and say: "you never have to change hardware.  If a
better  design becomes available,  just change  the software,  and the
atoms  will rearrange  themselves, and  presto  you would  have a  new
'hardware'  in place".   In  such a  situation,  the computer  program
itself could  transform the computer,  and this is certainly  the next
logical  step  in electronic  design  automation.   A  large chunk  of
hardware may then stand excluded from patenability, but then, it would
be a case of damnum sine injuria for hardware manufacturers.

More information about the Fsf-friends mailing list