WHEN THE Soviet Army newspaper published the results of the spring draft last month, the numbers did more than indicate deep trouble for the armed forces. They also revealed the underlying nationality conflicts and population differentials in various regions..

The report in Krasnaya Zvezda (Red Star) was the first to show a clear link between nationality questions and the draft -- in particular, a lack of response to the draft by non-Russians in the Soviet Union. There were great variations among the various republics and, in the case of Armenia, a startling deviation: In fact, the number of conscripts reporting for the May-June draft in Armenia totaled just 7.5 percent of those expected.

In Armenia, according to the January 1989 census, Russians constituted only 1.6 percent of the republic's population, with Armenians totaling 93.3 percent. Thus, the low turnout of conscripts could be interpreted as a turnout limited mostly to the sons of Russian and other resident nationalities. The figures also suggest that local problems, as well as changes in attitude toward neighboring republics and the dominant Russian state, led the Armenians to turn around completely. In 1989, by contrast, the ratio of those reporting for duty was said to be 100 percent of the expected number. In Georgia, where 6.3 percent are of Russian nationality, just 27.5 percent responded to the draft compared to a reported 94 percent last year.

Similar patterns are reported in the Baltic republics. And if the figures are not quite so low as in Armenia, the explanation may be the same: a military sign-up consisting of young Russian male residents, joined by those from several other nationalities and a few 18-year-olds of local nationality.

Following this pattern, only 40.2 percent of youths in Estonia reported for duty (down from 79.5 percent in 1989). Analogous figures for Latvia and Lithuania are 54.2 percent reporting in Latvia (down from 90.7 percent in 1989) and 33.6 percent in Lithuania (down from 91.6 percent in 1989). Russians comprise only 9.4 percent of the Lithuanian population. In Estonia, it is 30.3 percent, and in Latvia, 34 percent.

Radio Liberty monitored the Lithuanian capital's radio on July 25 and learned that 5,880 youths should have been drafted during the spring call-up. Of these, 4,704 were to be ethnic Lithuanians. However, only 960 of the latter showed up (20 percent) and about 1,400 non-Lithuanian 18-year-old males for a total of about 2,400. An attempt had been made to recruit others and the draft period had been extended to July 13.

But even with the extension, only 6 to 7 percent more (about 235) young men reported for duty -- still far short of the requirement. To the degree that Lithuania is typical of the other republics with larger proportions of Russians than Armenia, this pattern may well be duplicated as the drive for independence -- and an independent military -- gains among the individual non-Russian entities. There are some mysteries in the draft statistics. Why Moldavia, now the self-proclaimed Moldova, and Azerbaijan, with its internal security problems in Baku and Nagorno-Karabakh and problems with the Soviet military, are reported to have 100 percent prompt reporting for duty is inexplicable. Perhaps someone on the local level felt it necessary to show perfect behavior. In Azerbaijan, a higher share reporting in 1990 than 1989 makes the veracity of these figures for the republic even less likely to be correct.

Fascinating also are the drops in the other republics with large populations of Muslim origin, i.e., Kazakhstan and the four republics of Central Asia (Kirghizia, Tajikistan, Turkmenistan and Uzbekistan). In all cases, they declined from 100 percent last year to about 90 percent.

Given the declarations of sovereignty in nine republics, including frequent calls for their own armed forces, conversion of the Soviet military from a draft-based armed forces to a volunteer armed forces becomes more likely. A major step in this direction came this year when naval seamen were offered contracts as hired employees. In addition, the length of service of Navy draftees was reduced from three to two years.

Other factors are contributing to the desire by individual republics to have their own armed forces. According to Moscow News, some 15,000 soldiers have died from non-combat-related causes during the last four years. Of these, 3,900 perished last year. Other information indicates that some 70 percent of these "incidents" were nationality-related -- hazing, bullying, male rape and so forth. Uzbekistan alone witnessed the return of 510 sealed zinc coffins in 1989. (Uzbeks undoubtedly comprise the largest share of all conscriptees coming from the seven southern republics of Central Asia and the Transcaucasus.) Draftees from this same southern tier grew in proportional share of all draftees from 28 percent in 1980 up to a remarkable 37 percent by l988. If this trend were to continue until the end of the century, this region would contribute some 50 percent of all conscriptees. Earlier this month, an interview with the head of the Soviet military's Social and Psychological Problems Research Center revealed that the proportion of young males who "did not wish to serve" increased dramatically. Almost a decade before the Soviet incursion into Afghanistan, just 1.1 percent of those drafted in 1971 indicated that they did not wish to serve -- very likely a minimal rate who were honest in their response. By 1974, the rate of negative responses increased slightly, to 2 percent, and in the year of the invasion, 1979, 7 percent so responded.

However, by the time of the surveys conducted in 1989 and this year, the proportion of those "who do not wish to serve" increased to 18 percent, and of these, the conscriptees who came from large cities are cited as being much higher: 47 percent from Moscow, 38 percent in or from Leningrad, about 20 percent from Frunze. If this is consistent throughout the country, it suggests that rural males are more willing to serve.

But rural males more and more are from the demographically-driven pool of youths residing in rural Central Asia. And in addition to the reasons noted earlier for their unwillingness to serve, their Russian-language ability is said to be almost nil. From the Soviet military's point of view, it is a worrisome fact that among the 1.5 to 2 million draftees called up during 1989, some 125,000 spoke Russian poorly -- about one-and-one-half times greater than the year before and 12 times greater than in 1969.

How all of this affects our assessment of Soviet military capabilities remains to be seen. But nationality-related issues, combined with demographic shifts, health problems and a lack of Russian-language capability seem already to be changing the Soviets' view of themselves. Senior military officials there are now saying "not at the present time" rather than "never" for the institution of a professional armed service. Murray Feshbach is research professor of demography at Georgetown University. He has specialized in Soviet issues for three decades.