The Monitor's July 7 editorial entitled "The Ocalan Opportunity" calls for certain clarifications on the part of the Turkish Embassy. What you refer to as the "Kurd minority" should be placed in proper perspective. The composition of the Turkish nation is quite similar to that of the American nation -made up of many ethnic groups, not just Turks and Kurds. More important, these groups have been living together in peace for more than 1,000 years.
Turkish citizens of Kurdish origin are entitled to speak, publish, and broadcast in their native tongue. However, there is no uniform "Kurdish language" used by all ethnic Kurds in Turkey. The various dialects of Kurdish make it impossible for the respective members of the groups to communicate with each other. Therefore, Turkish is, and will remain, the common denominator for our entire nation.
When it comes to education, it is true that the Turkish state provides education solely in Turkish. But the same is valid for the French language in France. I should add that there is an ongoing debate in the US on the merits of providing bilingual education.
As for who speaks on behalf of Kurds, the greatest mistake would be to equate the PKK, a terrorist organization whose ruthless campaign has claimed thousands of innocent Kurdish and Turkish lives, with Turkish citizens of Kurdish origin. The overwhelming majority of our compatriots who come from a Kurdish background are loyal citizens of Turkey. They freely participate in the elections and send their representatives to the Parliament. In contrast, the PKK's active supporters are just a few thousand in number.
I hope you will consider these facts in your future assessment of political developments in Turkey.
Baki Ilkin Washington
Toward a better census
Allan J. Lichtman's and J. Gerald Hebert's June 22 opinion article ("The theory of counting heads") on using statistical methods as part of the decennial census missed the essence of the problem. Furthermore, I do not believe that there is a political conspiracy to purposefully undercount blacks and other minorities, as they suggest.
The principal problem is how to deal with non-respondents. The number of people who responded to the initial census mailing declined from 78 percent in 1970 to just 65 percent in 1990. This high rate of non-respondency caught the Bureau of the Census by surprise, adding significantly to the cost of the census, making 1990s count the most expensive ever at $2.6 billion. Some households had to be visited up to six times to secure the census survey. After adjusting for inflation, the overall cost per household in 1990 was 2.5 times higher than it was in 1960.
Statistical sampling could help in two ways. First, it is more cost effective. Second, it may provide an estimate on those who refuse or neglect to respond or cannot be found. But this second point raises an important issue. As anyone who is familiar with statistics knows, or ought to know, it is not possible to make a statistical inference from a sample without having the distinct possibility of introducing an error.
When enumeration is not possible, for practical or cost reasons, then statistical methods are the next best thing.
Certainly we may have some small segments of society who refuse to cooperate with the census, and, perhaps, some statistical sampling might provide a better number. That number, however, would likely be less accurate than if the bureau improves its enumeration methodology, which I believe it has done, and conducts an aggressive public information campaign in order to overcome the barriers to participation.
Erik D. Randolph Penbrook, Pa.
The Monitor welcomes your letters and opinion articles. Due to volume, we can neither acknowledge nor return unpublished submissions. Letters must be signed and include your mailing address and telephone number. Mail letters to 'Readers Write,' and opinion articles to Opinion Page, One Norway St., Boston, MA 02115, or fax to 617-450-2317, or e-mail to email@example.com