[Paper presented at the Equity Diversity and Information Technology (EDIT) at National Institute of Advance Studies (NIAS), Bangalore in December 1999]
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
Abstract
Different organizations have been working to make the computers available in the different Indian languages. However, because of multiplicity of the languages (there are 18 languages recognized by the Constitution of India) the issue is quite complicated. There are only two essential components required to represent a language on computers - The language must have a Script, and it should be possible to represent the script on the computers. The computers understand English because they were developed by people who used English. We have to understand and accept that the fault why computers do not understand our language does not lie in our language. Bharatbhasha, a "not-for-profit" organisation, supports 5 languages -Hindi, Marathi, Gujarati, Gurmukhi and Bangla. Bharatbhasha uses a different system for mapping the keys on the keyboard to the Indian language alphabets. This system is called "Roman-Indian" keyboard system. It uses the English keyboard for mapping Hindi and other Indian Language alphabets. Most of the Indian programmers and computer users know English and one of the Indian languages and find this system very simple to understand, learn and use. This system, for most part is intuitive to people knowing English and Indian languages (Marathi, Hindi, Gujarati, Gurmukhi & Bangla). If one is familiar with Standard English keyboard and knows Indian languages (Marathi, Hindi, Gujarati, Gurmukhi & Bangla) one will take only 20/30 minutes to learn the Indian language keyboard. This system uses the fact that for most of the characters in one language there are equivalent characters in most of the other Indian languages. Bharatbhasha system had shown that one could use computers in Indian languages without paying the extra cost for the hardware and software which the other solutions required. This system benefits both those who want to use Computers in Indian Languages and those who prepare software in Indian languages since it is free. The market for Indian Language software is quite large. Can the industry ignore this market till the market learns English? Computer applications in Indian languages can be prepared by the people who know computer programming, therefore, this part of the work has to be taken up by those who know computers and also know Indian languages. Others can and must help in standardization, development of terminology in the local languages, developing specification for software and negotiating with the Banks and the Government for softer loans and reduced taxes etc.. However, they all must work in unison.
Keywords: Localization, Bharatbhasha, Indian Languages, Hindi, Marathi, Gujarati, Gurmukhi, Bangla, Computers, India and Standardization
Top of the page
Back to Articles page
Mail me your feedback or any question you have on this.
I am sure you all have read enough about change. Anybody who is somebody has written about change, pace of change and its impact on our lives and society. Some of the names that come to my mind are Peter Drucker, Alvin Tofler, Naisbitt, Negroponte and of course, Bill Gates. While change is a natural way of life, the change these people have talked about is the quantum change brought about by new technology. This type of change may be new to us but certainly not new to our society, which has seen many technological innovations in past and the changes effected by them. When we study or read about them we find that each one of these changes or "revolution" did not affect all sections of society equally. Printing press by Gutenberg shifted the work from the monks to the printing press, brought the cost lower and facilitated delivery of technology benefits to a wider population. Invention of Wireless by Marconi shifted the work from the delivery boys to the telegraph companies but increased the speed, reduced the costs and delivered the benefits to a wider population. Invention of Steam Engine and Motor Car wiped out the horse carriage but reduced the costs of transportation and gave benefits to trade and industry and society at large. I am sure that those of the horse carriage owners who changed and took to the new trade/technology must have survived and flourished while others must have been marginalised. I am stating these as a matter of fact and not being either optimistic or pessimistic about the change.
I only submit that we have to accept change as a way of life and see what we can and must do to help all sections of society to reap benefits of technology. This responsibility is also perhaps not new and must have existed in all ages and groups like ours must have discussed, debated and implemented schemes to help the disadvantaged sections of society. But our responsibility is perhaps higher because of the peculiarities of this information/knowledge revolution. This revolution is different from the earlier ones. In the earlier revolutions the prime resource one needed to flourish was land or capital and these resources were not equitably distributed among the different sections of society. In the current revolution the prime resource one requires to flourish is the human brain, which God has been kind enough to distribute to different sections of society in equitable manner and, therefore, our responsibility is higher than that of similar groups in past. But we must remember that an uneducated person in this revolution is like the unskilled/disabled person in the earlier revolutions.
However, the other prime resource required in this revolution is "Information", which is neither distributed in equitable manner nor is it being produced in equitable manner. These inequities in production and distribution of information among the countries have been discussed by Mr. S Arunachalam (Endnote 1), who concludes by saying "By every yardstick one can think of, the developing countries lag far behind the advanced countries in the enterprise of knowledge production". Inequities exist not only between the nations but also within a nation between the different sections of society. I also believe that if steps are not taken then the gap between the rich and poor will further increase (Endnote 2).
In this age the Information and Communication Technology (ICT) tools will play a crucial role and the language in which these tools work "...is a key variable in determining who benefits, who gains, who is excluded, who is included and how the "Information Age" impacts the cultures of the world" (Endnote 3). Bangalore Declaration (Endnote 4) has also raised a note of caution on this issue in clause 26 b, which states - "Even within countries dominance of the English speaking urban elite is already nationalizing indigenous cultures. IT will worsen this situation since it is the same elite that has historically enjoyed disproportionate benefits from any new technology, to the disadvantage of the weaker section of society." I feel that despite the efforts to further the cause of Indian languages on computers Digirati (Endnote 5) may well emerge as the new ruling class even in our country.
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
Though Bish Sanlay and Donald A Schon (Endnote 6) have written for a different context, a large part (excluding the parts specific to the socio-economic, legal and political system of the US) of their article is very relevant to Indian context also. In their context one divides the population into two groups – the rich and the poor. However, the problem in Indian context is more complex because the computers, at present, cater to the English knowing population only. There are attempts by different organizations to make the computers available in the different Indian languages. However, multiplicity of the languages further compounds the problem.
Though there are "1652 mother tongues listed in census, 33 are spoken by people numbering over a lakh" (Endnote 7). Eighteen of these languages are recognized in the 8th schedule of the Constitution of India, which has one full part (Part XVII) devoted to "Official Language". This schedule originally had 15 languages and three languages (Konakani Manipuri and Nepali) were added in 1992 by the 71st amendment of the constitution. Oldest of these languages is perhaps Sanskrit (between 2000 and 500 BC). Most of these have their own script but some of them are written in more than one script- Sindhi is written in Urdu/Arabic/Devnagiri; Konkani is spoken in parts of Maharashtra, Karnataka, Goa and Kerala and is written in more that one script. Article 343 of the Constitution declares Hindi as the "Official Language" of the Union- "The official language of the Union shall be Hindi in Devnagiri script. The form of numerals to be used for the official purpose of the Union shall be the international form of Indian numerals."
However, it also allows use of English and Devnagiri form of numerals till such time as the "Parliament may by law provide". It also permits the Legislature of a state to adopt one or more language for the official purpose of the state (Article 345). Consequently most of the states have adopted state languages. There are a number of rules (Endnote 8) about language to be adopted by the states in correspondence with each other. Schools and colleges in most of the states teach two or three languages. There is also no uniformity in the medium of education adopted by schools and colleges. Higher technical education is delivered in English while most Primary and Middle level schools and colleges use Hindi or the state language as the medium of education.
Whenever we consider/assess size of the Indian market for any language sensitive Information and Communication Technology (ICT) application (e.g., Pager, Computer, Internet etc.) we must consider the language spoken by the market segment as an important variable. We must also keep in mind the multiplicity of languages and the fact that a large number of people do not understand English. Therefore, whenever we consider the market segments for "Language Sensitive/Non-Language Neutral" technology we may represent the Indian situation as in the matrix below:
Language Matrix
| 1. English Knowing
(5%) |
2. Not-Knowing English
(95%) | |
|
1. Poor |
Very small group |
Very large group |
|
2. Middle Class |
Large group |
Substantial Number-many times the English-knowing group. |
|
3. Rich |
Large group |
Small group |
Table 1 - Language Matrix
This, however, is rather misleading representation because the relative sizes of the segments does not come out clearly in this. Following is, perhaps, a better representation:
 |
As this figure indicates the population in the middle group and not knowing English is many times the English knowing population in the same group. This group, besides the "Rich" group certainly has purchasing power for the ICT applications. |
| Figure 1 |
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
Before I mention about our experience of Bharatbhasha let me explain what is required for representing any language on computers. There are only two essential requirements to represent a language on computers:
- The language must have a Script, and
- There must be a way to represent/show the script on the computers.
Let me remind you that the computers do not understand any languages other than "0" and "1". All codes/words/programs are finally translated into these two digits before being processed in the computer (Figure 2).
 |
 |
|
Figure 2 |
Figure 3 |
However, since humans are not very comfortable in using just these two symbols and they use a language to communicate with each other they have taught their language to computers. For example in the picture on the right hand side (Figure 3) it is shown that to copy one file "file.doc" from drive "a" to drive "b" we write "copy a:\file.doc b:\" [common DOS command]. The code to which word/command "copy" refers accepts two arguments and copies the file name passed by the first argument to the location passed through the second argument.
Figure 4 |
It does not matter whether we call the command "copy", "COPY" "Kapi" or even just "xyz" or for that matter word in "Fig. 4", which is the equivalent word in Hindi. This will work as long as the code referred to by it is the same and the two required arguments are passed on. We must understand that it is our decision as to what we want to call this command (code) and no matter what we call it will do what it as been programmed to do. It is mentioned that codes for all the "reserved" words for every computer language are available and all one has to do to port a language is to change the "labels" by which these codes are addressed.
The computers understand English because they were developed by people who used English. Some of the other countries, having economic power or number power demanded and had the computers in their languages. A number of other countries too have computers in their language, e.g., China, Japan, Germany, France etc.. Some countries like Iceland tried to have the computers in their language but failed (Endnote 9). Unfortunately we in India did/could not demand and have the computers that understand our language. Only recently Microsoft has announced that the Windows 2000 will understand some of the Indian Languages. We have to understand and accept that the fault why computers do not understand our language does not lie in our language. Our languages have a script and there are applications available to represent them on computers and, therefore, computers can and must understand them. But that, as we will see, only meets the necessary but not the sufficient condition. We have only ourselves to blame why the computers do not understand our language.
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
Bharatbhasha (Endnote 10) is a "not-for-profit" organization started in April 1997 when Mr. Venkatesh Hariharan joined our group, which till then had consisted of only two members, i.e., my wife and me. The group has grown since then and is quite big now. We at Bharatbhasha have tried to address to the problems of the "Not-Knowing English" group [Table 1 and Figure 1] by providing means to facilitate development of packages and applications in Indian languages. As on date 5 languages are supported – Hindi, Marathi, Gujarati, Gurmukhi and Bangla. Let me take you through the journey we have traveled.
Before I start discussing the work done by Bharatbhasha, I must mention some of the peculiarities of scripts of Indian languages. The script of most of the Indian languages is such that you cannot write it in the four lines in which we write English. We require one additional line as explained in the figures "5" & "6".
 |
 |
|
Figure 5 |
Figure 6 |
Vowels are joined with the consonants in the form of "Matrian" which come either after (arrow 1 & 3 below), before (arrow 2 below), on top (arrow 4 below) or below (arrow 1 below) the consonant [figure 7].
 |
|
Figure 7 |
There is also a concept of "half" and "full" character. These are shown in the figure 8 below. The half of character "b" in figure 8 is written as in "c" in figure 8 below if it comes in the end of the word else it is written as in "a" in figure 8 below. Full character is shown in "b" in figure 8. The purpose of this article and space does not permit me to go into the other intricacies of the Indian Scripts and those interested may refer to the article by Dr. Mudur (Endnote 11).
 | ||
|
a |
b |
c |
|
Figure 8 | ||
In addition to these peculiarities there are 3 or 4 different types of key boards (CDAC, DOE, Remington and Godrej etc.) in Hindi and other Indian languages. The differences between these keyboards are such that a person trained on one type of keyboard will find typing on another type of keyboard very difficult, if not impossible. One of these keyboards is given in figure 9.
|
|
| Figure 9- One of the Hindi Keyboards |
Most of the people knowing English keyboard find it very difficult (nearly impossible) to remember this keyboard and find it even harder to use it. One has to be familiar with the keyboard to be able to type on a computer. It is not very difficult to remember one keyboard layout and most people working on computers in India are familiar with the English keyboard. Problem comes when they have to remember two keyboards -one to type in English and another to type in Hindi/Indian Language. Remembering two keyboards is really difficult and confusing for most of the users. As a result of this people working on computers often keep working only in English or at best they manage by typing in "Roman-Hindi", i.e., Hindi or the other Indian Languages in Roman Script. This happens even in Government Departments where the rules (Endnote 12) often require that the letters are sent in Rajbhasha (the "Official Indian Language") and/or in the language of the state besides English. They get around to this by passing on this work to the "Rajbhasha" section instead of learning to type in Indian language and doing this work themselves.
Bharatbhasha uses a different system for mapping the keys on the keyboard to the Indian language alphabets. This system is called "Roman-Indian" keyboard system. It has been designed by taking components of Phonetic, Roman Hindi and Graphic system. It uses the English keyboard for mapping Hindi alphabets. Hindi characters are mapped to their corresponding English equivalents/near equivalents (figure 10).
As a result a major part of the layout is obvious to most people who know Hindi and are familiar with English keyboard. They take very little time to learn this keyboard as they do not have to learn two different keyboards. The suggested system, for most part is intuitive to people knowing English and they have to remember only a few keys. The keyboard used is very much the Standard English keyboard and normally no stickers need be fixed. Most of the Indian programmers and computer users know English and one of the Indian languages and they find this system very simple to understand, learn and use. If one is familiar with Standard English keyboard and knows Indian languages (Marathi, Hindi, Gujarati, Gurmukhi & Bangla) one will take only 20/30 minutes to learn the Indian language keyboard. The system for Hindi can be seen from the detailed key mappings in figure 10 below and for other languages can be seen from Appendix 1:
 |
| Figure 10- Bharatbhasha keyboard for Hindi |
Let me explain this system in more details:
In this system all the components of the script have been provided and therefore all the characters / symbols of the script can be generated/created. The system can also be used to write the web pages also by selecting the tags <FONT> & </FONT> as sown below:
<FONT SIZE=5 FACE="Shusha">AaAao ihndI AaOr marazI maoM ilaKoM.</FONT>
The above text on the web page will read as "AaAao ihndI AaOr marazI maoM ilaKoM" if the required fonts are not available else they will read in Hindi as in figure 11.
Figure 11
This system, for most part is intuitive to people knowing English and Indian languages (Marathi, Hindi, Gujarati, Gurmukhi & Bangla) as can be seen from the detailed key mappings. Detailed mapping of the 256 characters is enclosed in "Appendix 2". A number of ASCII values have not been used to so that they can later cater to the characters of the other scripts as and when the system is expanded to include them.
Let me stress one point that I had made earlier – the computer primarily deals with numbers only. For example the word "kumar" is kept as the ASCII values 107, 117,109, 97, 114 and will be shown as "kumar", "kumar", "kumar" or "kumar" depending on whether we choose the fonts as Arial, Emboss, Lucida Handwriting or Serifa BT. The font file merely supplies the picture/shape stored at that ASCII value and the characters/words thus formed get shown on the screen. We have used same principle. For example if we type "kumaar" we will see the equivalent word of "Kumar" in Hindi, Marathi and Gujarati depending on whether we have selected "Shusha", "Shivaji01" or "Vakil_01" as the fonts (Figure 12). For typing this word in Gurmukhi and Bangla we have to type only "kumar" and select the fonts as "Sandhu01" or "Shymal01".
 |
| Hindi | Marathi | Gujarati | Gurmukhi | Bangla |
| Figure 12 – Word "Kumar" in the different Indian Languages | ||||
A little difference in typing in these languages is because there are no "Half" characters in Gurmukhi and Bangla. The fact that some of the Indian Languages have Half and full characters was mentioned in Figure 7 above. This also shows the ease with which one can learn to type in the different Indian languages. In fact if one knows English Keyboard and knows writing in one of the Indian languages one can type in any of the other supported Indian languages. I must mention a story. When the daughter of my Gujarati guru went to USA for higher studies she had a problem in communicating with her grandmother because while she could speak Guarati she did not know how to write in Gujarati and the Grandmother could only read Gujarati. She solved the problem by using this system. She used to type Gujarati in Hindi script in "Shusha" and then change the fonts to "Vakil_01" and E-Mail the letter to her father who used to take a printout and give the Grandmother. This was possible because this system uses similarity between the scripts of the two languages and consequent similarity between the key mappings as can be seen from Appendix 1.
There is so much similarity in the keyboards for the different languages that one has to learn only one keyboard and one can then type in all the other languages also. This system uses the fact that for most of the characters in one language there are equivalent characters in most of the other Indian languages. We have mapped these in such a manner that the character equivalent to (say) "A" (ASCII – 65) is always at the same ASCII value for all the scripts – Hindi, Marathi, Gujarati, Gurmukhi and Bangla. This is explained in the figure 13, where L1, L2, L3, L4 & L5 are the different languages- Hindi, Gujarati, Gurmukhi, Bangla and English. It may be seen that in our system character equivalent of "A" in all these languages is mapped to the ASCII value of character "A", i.e., 65. Similar mapping is done for the other characters also as can be seen in the figure 13. Those interested in exploring the similarities between the different Indian Scripts may like to refer to Ganesh-Vidya (Endnote 13) a very fascinating book I came across recently. I wish that I had got this book three years back.
Figure 13 – Character Mappings
The other strong point of this system is that it does not use any software/program to generate the keyboard. Since it is based on the true type fonts and does not use any TSR (Terminate and Stay Resident) program or DLL (Dynamic Link Libraries) it can be used to do whatever you want to do in the supported Indian languages-Typing, preparing Indian language front-ends, Web pages, sending E-Mails etc. etc.. It can be used in all the object oriented development tool which allow "captions" of objects to be defined in different fonts. In all such tools we can write the user interface in Hindi and other supported Indian languages. In other words using this system you can write programs & packages in Visual C++, Visual Basic, Visual FoxPro etc., which have look and feel of windows and messages, menus etc. in Indian languages and user does not need to learn English just to use the computer/software. This system, therefore, enables the programmers to write programs, which have more user friendly front ends in local languages. A detailed "Howto" and the source code of a sample multilingual application can be seen at the site:
This has always been distributed for FREE through the direct contact, through the net, floppies, e-mails and CDs with the popular Indian IT magazines. We have thus distributed about 75,000 copies of this system to different people since March 1997.
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
 |
At present there are different coding systems used by the different vendors. These coding schemes are different from each other in such a manner that the files typed in one package/application are not readable in other application. This is explained as in figure 14. L1 is the Bharatbhasha coding scheme and L5 is the English coding scheme; V-1, V-2 & V-3 are the schemes used by three different vendors. It may be seen that the character equivalent of "A" is stored at the ASCII value of 180, 163, 123 by the three different vendors. As a result text typed in one package will appear garbled in the other package. One has to write a "Filter" which will correctly |
|
Figure 14 |
substitute character from one vendor specific |
| system to another vendor specific system. However, writing filters is not an easy task because different vendors do not provide for same set of characters/script elements. |
At present there are different coding systems used by the different vendors. These coding schemes are different from each other in such a manner that the files typed in one package/application are not readable in other application. This is explained as in figure 14. L1 is the Bharatbhasha coding scheme and L5 is the English coding scheme; V-1, V-2 & V-3 are the schemes used by three different vendors. It may be seen that the character equivalent of "A" is stored at the ASCII value of 180, 163, 123 by the three different vendors. As a result text typed in one package will appear garbled in the other package. One has to write a "Filter" which will correctly substitute character from one vendor specific system to another vendor specific system. However, writing filters is not an easy task because different vendors do not provide for same set of characters/script elements.
Older standard, namely ISCII, is not usable in windows, as it does not provide for the half characters explained in the figure 7 above. There are a couple of Government bodies that have been working on Indian language software but unfortunately they do not see eye to eye with each other on the important issue of standardization. I do not wish to name them. Professor Keniston (Endnote 14) has correctly identified and rightly pointed out that there is a basic difference in the approach adopted by these two organizations and consequently the differences are, perhaps, irreconcilable if these organizations are left to themselves. They are both trying to solve this problem through the market. They have been furthering their own systems and hoping that their system will come out winner in the market. Let us see what happens. It is however, certain that both will not come out as the winners and of course the end user will certainly be the loser.
When we started in 1997 and we had two options. One was to do nothing and wait, hoping that these giants will some day resolve the differences and come up with a commonly acceptable standard; the other option was to make one open standard and to follow it and to switch as and when these giants resolved their difference. We followed the second option.
Fortunately there has been some change in the scenario. The Government of Tamil Nadu (Endnote 15), called a meeting to sort out this issue for Tamil. They had called the Computer professionals, University Professors from Computer Science and Linguistics Departments, Government Bodies, Merchant Chambers and all the other stakeholders. They have successfully resolved this problem for their state language, i.e., Tamil. We were very happy with this meeting primarily because this was the first time a state Government had recognized the problem and had tried to solve it by taking command over the situation and forcing all the stakeholders to assemble on the negotiating table and arriving at consensus. Secondly, we were happy because one of the keyboards standardized by them used "Roman Indian" system used by Bharatbhasha, discussed earlier in his article. Attention to this issue has also been drawn in the clause 30 b of the Bangalore Declaration. I wish that other state governments also take initiative and sort out this issue.
We hear that Windows 2000 will understand some of the Indian Languages and I am afraid that if the State Governments do not sort out the issues it may be too late for them to do any thing, for then Microsoft will decide what will be the standard. Without meaning any offence to Microsoft, I feel that the present situation is similar to the one that occurred about 1000 years back when Indian rulers kept quarrelling among themselves and one of them invited Mahmud of Gaur to settle scores with others. I only hope what happened about 1000 years back in history does not get repeated and we do not suffer because the giants do not discuss with each other across the table.
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
Demand for Indian Language software? Will it sell?
When we started it was never so easy and cheap to prepare software in local languages. At that time very few software were available in Indian languages and that too mainly for word processing applications. All these software required money and some of these even required additional hardware. It is important to point out that earlier (a year or two back) if one wanted to use computers in Indian languages one would have had to spend about Rs. 8,000/- to Rs. 10,000/- more than the one who did not want to use computers in Indian languages. In other words there was a hidden penalty to be paid to work in Indian Languages. No wonder most people used it only to work in English and computing in Indian Languages had not picked up. Bharatbhasha system had shown that one could use computers in Indian languages without paying this extra cost. This system benefits both those who want to use Computers in Indian Languages and those who prepare software in Indian languages since it is free and they have to carry no additional burden.
Our approach has been that Fonts for Windows programmers are like Oxygen is to living organisms. These are the very basic needs, and therefore, should be available for free. Any one can use these and write applications and earn money by charging for the software and not for the fonts.
Let us now consider the question of demand for such software. Let me tell you a story. Once two salesmen went to a country and found that people of that country did not use footwear. One of them reported to his head office- "There is no demand for footwear in this country as people here do not use any footwear". The other reported – "There is tremendous demand for footwear in this country as people here do not use any footwear". I leave it to you to decide which salesman was right.
At present computerization efforts and benefits both have remained limited to only the English knowing population, i.e., about 5% of the total population. Computerization has not touched the larger masses. Someone could argue about the purchasing powers of the larger masses and other issues. Let us consider one section of this large mass. If you look at the work of small shopkeepers in any town - Mumbai, Badaun, Gaya, Trichy, Bhavnagar or any other -you will find that there is no reason to doubt their purchasing power or to believe that they will not be able to benefit from computerization. They too can benefit from computerization provided there is a product available to them and they can adopt and implement it successfully. This product should be able to let them do their work in their language. At present their problem in computerization has been that they have to learn English to use computers or employ an English knowing employee for this work. Even employing another hand is not a problem but the problem is that they have to share their closely guarded business secrets with this person. They would certainly be very comfortable with computerization if there is a product which they could use in the language they know. A little calculation of Return on Investment in Computerization will tell that computerization of Inventory management and preparation of Accounting Returns at these small shops will pay for itself in 6 to 8 months (saving about 2 hours of time each day, i.e., 14 hours a week). Recently one of the Indian IT magazines had reported that the sharpest rise in the sale of PCs had been in the smaller towns. I have also spoken to a large number of people in this group in towns like Mumbai, Vadodara, Delhi, Gurgaon, Jaipur, Ajmer etc. and have received mail (both postal and electronic) from places like Badaun, Kashi, Belgaon, Jalgaon etc. etc. I find that there is plenty of scope to sell Indian Language software atleast to the shopkeepers.
IT industry in India has to think about their approach to deal with this large market (having purchase power but not knowing English). Can the industry ignore this market till the market learns English? In which case they may have to wait for a generation and hope that the next generation of the members of this market will know sufficient English to use computers. Incidentally the size of this market is estimated to be about 20 to 30 times the size of the existing English knowing market.
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
Who should/will work for Indian Language software?
This is perhaps the most important question. Preparation of Computer programs in Indian languages can be done by the people who know computer programming, therefore, this part of the work has to be taken up by those who know computers and also know Indian languages. We have a number of programmers in India and for most of them English has been second or the third language. They will be able to prepare software with local language front-ends. This group will need to coin new words in Indian languages for the different terms and menu items. Those who understand the languages and are monitoring/studying its development are perhaps best suited for this work. These groups have to be supported by the local bodies of the markets and organizations, which promote use of local languages, e.g., different Sahitya Academies and the "Pracharni Sabhas" etc.. The local bodies of the markets may help in defining the software specifications and the language bodies will help in developing terminology and in promoting its use. All these will have to talk to the Government bodies and also negotiate with banks etc. for cheaper loans. The Government Bodies may have to be asked for reduction or exemption of Excise duty, Sales Tax and other duties and taxes on such software and Hardware. These groups could also talk to the government for exempting income from such software from Income Tax at-least for the next 5 years or so. Some of these aspects have also been covered in Clause 13 b of the Bangalore Declaration (Endnote 16).
I do not wish to give an impression that the big computer companies do not have a role to play in this. Their role will be very important for they will face a big threat and will have to fight for market share. This is a new market and each of them would like to have a chunk of this market. Each of them will have to chalk out their strategy for this new market. A study of customer list of big computer companies reveals that they are very comfortable in handling large accounts. Small accounts somehow do not get much attention from them. Small customers are not found attractive by them as they are thinly spread over larger area and perhaps it becomes difficult for these companies to service them. Their big accounts have about 50-100 computers spread over 3 or 4 sites in a city like Mumbai. In other words even the large accounts have only about 20 or 30 computers in one area or building. The density does not rise as the larger organizations have their offices spread over a large area in big towns. This market is very different from the new market I have been talking about for software in local language.
In the new market, of the size of (say) Vashi (Endnote 17), there will be about 400 computers in a stretch of less than a kilometer. Supporting and servicing such concentration will be economically viable for any software or hardware vendor. Such large number of computers in such small area will be far easier to maintain than maintaining even the largest accounts of today and therefore, these will be far more attractive than today’s large accounts. Full market potential can be imagined when we see that there will be hundreds of such dense clusters - Dadar, Parel, Crafford Market and similar markets in small towns spread all over the country. Big companies, if they want to remain big, will have to learn to deal with this new market where there would be a large number of small customers. Next wave of computerization in India will belong to the company which can cater to and capture this large market.
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
This is perhaps a very difficult question for we do not know where to begin. There is so much of work remaining to be done. We would be better off if we do not treat it as a linear process and waste time in arguing about what comes first and what next. There are a number of issues that require immediate attention. I am listing them down as they come to my mind and not in any predefined order.
- The Central Government of India has recently formed a separate ministry for Information Technology (IT). Most of the IT related departments which were earlier under the different ministries have now been put under this new ministry. I am sure it will improve coordination between them and help them work in unison. The department of Official Languages has to work harder faster to develop new terminology in the local languages.
- The state governments have to take lead for standardization codes and terminology in their respective language. They have to force all the stakeholders to come together and decide/finalise the standards and keep them "Open". They will also have to arrange free availability of the fonts and libraries of the standard terminology.
- Both the Central and the State Governments have to think of ways (simplifications of rules and procedures, giving tax concessions, influencing/formulating schemes and policies of the financial institutions for arranging venture capital or softer loans etc.) to facilitate development of software in Indian languages. They have also to ask the IT task force of the Government of India and state governments to apply their mind to this issue/problem, find and implement solutions and also to monitor progress made in this direction.
- Language bodies have to start developing the terminology and to publish them in a manner that they are freely available to everyone. They will have to make a compromise between the generally acceptable words and the pedigreed "Sanskritized" words. These bodies will also have to remember that despite the best efforts of the Government of India to popularize the two words –"Doorbhash" for telephone and "Doordarshan" for Television these have not been accepted by the public. The word "Doordarshan" to people does not mean Television but it means the broadcast from the Government Television Center in New Delhi. The people have not accepted "Doorbhash" for telephone.
- Local talents should be used to develop contents in the local language whether it is for computer applications or for the web. This is an enormous task and has to be taken up by a number of bodies simultaneously. As long as the contents are not in the local language they have to be aimed at the intermediaries, like the Village Panchayat, Village workers, Health workers, Non-Government Organizations etc.. Work done by the MS Swaminathan Foundation (Endnote 18) in Pondicherry also suggests that focusing on the Panchayats helps – one PC and three volunteers trained in handling PC and data communication helped the whole village to reap the benefits.
- The Agricultural Universities and the Institutes of the Indian Council of Agricultural Research (ICAR) must take lead and setup depositories of knowledge/Information which can be accessed to help the large section of population which is engaged in agriculture.
- Develop an OS in Indian languages. Linux is free Operating System (OS) and can easily be used to make the Indian language OS. Fortunately, during the EDIT seminar at NIAS Bangalore, Dr. Sadgopan, Director of Indian Institute of Information Technology (IIIT), Bangalore quite happily agreed to take the lead and Prof Sangal of IIIT, Hyderabad agreed to extend all help. We hope to begin this work from January 2000 onwards.
At present we in India some times boast of our large computer trained manpower resource. This is an ever growing resource because of the growing population and the fact that we do not find any town in India which does not have a computer training center. All the boys and girls passing out of these will certainly not be able to get jobs in the US and other countries. We will have to find jobs for them in India. I cannot think of finding a better way of creating jobs for them than hoping that they will take to developing software in the local language and satisfy the needs of the local markets. This will not only help them in finding jobs but will also channelise benefits of IT to the non-English speaking group. I do sincerely wish and pray that Computing in Indian languages picks up otherwise after a few years we may have the dubious distinction of having the largest pool of unemployed IT trained manpower.
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.
Endnotes
- Arunachalam, S, Information and Knowledge in the Age of Electronic Revolution: A developing Country Perspective, Text of Talk delivered at the Volkswagen Foundation, May 7th 1999.
- Dertouzos, Michael, The Rich People’s Computer?, People’s Computer, January/February 1999
- Keniston, Kenneth, "Language, Power, and Software", Paper Prepared for Conference on Localization, Center for Development of Advanced Computing, Poona, Maharastra, India, September, 1998.
- The Bangalore Declaration on Information Technology for Developing Countries in the Global Village, National Institute of Advanced Studies, Bangalore, Nov. 4th 1998, ISBN–81-901089-5-6.
- Keniston, Kenneth, "Language, Power, and Software", Paper Presented at the Computer Society of Nepal Convention in Nepal in January 1998.
- Sanyal, Bish & Schon, Donald A., Information Technology and Urban Poverty: The Role of Public Policy, High Technology and Low-Income Communities, edited by Donald A. Schon, Bish Sanyal, William J. Mitchell
- Manorama yearbook 1999: Chief Editor KM Mathew ISSN 0542-5778, page 480; one lakh = 100,000
- Article 346 of the Constitution of India and the rules and procedures made under it particularly "The Languages Rules 1976". There is a manual called "Manual Regarding the use of the Official Language Hindi" second edition (revised upto 29th February 1996) issued by the Department of Official Language, Ministry of Home Affairs, Government of India which is a very good compendium of these rules and procedures.
- Welsh, Mary Williams, "Windows won’t compute into ancient Icelandic Language" in Los Angeles Times, Posted on June 30 1998.
- Bharatbhasha is a registered Charitable Public Trust.
- Mudur, SP, and others, An architecture for the shaping of the Indic Text, Computers and Graphics 23 (1999) 7-24.
- Endnote 8 above.
- Wakankar, LS, Ganesh-Vidya, The traditional Indian approach to Phonetic Writing, Published on he eve of the Xth Congress f the Association Typographique Internationale, Paris (6-10 November 1967) by R Subbu, Convenor, Script Study Group, Mumbai.
- Keniston, Kenneth, "Language, Power and Software", for Charles Ess, ed. Cultural Attitudes Towards and Communication, New Your: Suny Press, 1999.
- One of the important states of India, Chennai (Madras) is its capital. It is in the Southern pat of India. Tamil is the state language.
- Endnote 4 above.
- Vashi is a suburb of Mumbai on the Eastern side at a distance of about 35 KM from Chhatrapati Shivaji Terminus (CST) in Mumbai. It is a very important population center in the satellite town of New Mumbai.
- http://www.mssrf.org/information%20village/knowledge-system-program.html
Top of the page
Back to Articles page
Mail me your
feedback or any question you have on this.