Author's note: Research on this paper was partly carried out as a Peace Fellow of the United States Institute of Peace. The author would particularly like to thank Paula Smith, Mohammad Eshaq, Hafizullah Karzai, Professor Hasan Kakar, Anthony Arnold, David Katz, Houchang Chehabi, and five anonymous reviewers for their assistance and comments. All views are those of the author, not of the institute or anyone named above.

¹ Theda, Skocpol, States and Social Revolutions (Cambridge, 1979), p. 165.Google Scholar

² For a review of a variety of such studies, see Robert, D. Putnam, The Comparative Study of Political Elites (Englewood Cliffs, N.J., 1976), pp. 170–72.Google Scholar

³ Shahrani, M. Nazif, “Causes and Context of Responses to the Saur Revolution in Badakhshan,” in Shahrani, M. Nazif and Robert, L. Canfield, eds., Revolutions and Rebellions in Afghanistan: Anthropological Perspectives (Berkeley, 1984), p. 166; Shahrani borrows the notion of “fields of power”Google Scholar from Wolf, E., Peasant Wars of the Twentieth Century (New York, 1969), p. 290.Google Scholar

⁴ For overviews of state formation in Afghanistan, see Shahrani, M. Nazif, “State Building and Social Fragmentation in Afghanistan: A Historical Perspective,” in Ali, Banuazizi and Myron, Weiner, eds., The State, Religion, and Ethnic Politics: Afghanistan, Iran, and Pakistan (Syracuse, N.Y., 1986), pp. 23–74;Google ScholarAshraf, Ghani, “Afghanistan xi. Administration” in Ehsan, Yarshater, ed., Encyclopaedia Iranica, vol. I (London, 1982), pp. 558–64;Google ScholarBarnett, R. Rubin, “Lineages of the State in Afghanistan,” Asian Survey (11 1988), 1188–209.Google Scholar

⁵ Giacomo, Luciani, “Allocation vs. Production States: A Theoretical Framework,” in Hazem, Beblawi and Giacomo, Luciani, eds., Nation, State and Integration in the Arab World (London, 1987), vol. II,Google ScholarThe Rentier State, p. 69. Luciani suggests defining “allocation states as all those states whose revenue derives predominantly (more than 40 percent) from oil or other foreign sources and whose expenditure is a substantial share of GDP.” Afghanistan meets the first part of the definition but not the second. For data,Google Scholar see Rubin, , “Lineages of the State in Afghanistan,” Table 1, p. 1202.Google Scholar

⁶ For an analogous distinction, see Kirin, Aziz Chaudhry, “The Price of Wealth: Business and State in Labor Remittance and Oil Economies,” International Organization, 43 (1989), 105–9.Google Scholar

⁷ The Persian spoken in Afghanistan is officially known as Dari. Its degree of difference from Iranian Persian seems to this nonhinguist to be about the same as that of American from British English.

⁸ Non-Muslims, including tiny groups of Hindus, Sikhs, and Jews, played roles more akin to consulting firms (in trade and finance) than citizens. A small group of urban Shiʿites (Qizilbash) played a similar role in the bazaar and bureaucracy.

⁹ In Afghanistan the term “Kabuli” is more commonly an ethnic designation for members of the qawm (solidarity group) of the indigenous Persian speakers of Kabul. My term, “Kabuli Pashtuns,” would not be used by Afghans. While in Afghanistan the term “ Afghan” still commonly means “Pashtun,” I am using it in its juridical sense, meaning any citizen of the state or anything pertaining to the country.

¹⁰ Wolfram, Eberhard, “Afghanistan's Young Elite,” Asian Survey (02 1962), 10.Google Scholar

¹¹ Gilbert, Etienne, L'Afghanistan, ou les aléas de Ia coopération (Paris, 1972), p. 37.Google Scholar

¹² Barnett R. Rubin, “The Old Regime in Afghanistan: Recruitment and Training of a State Elite,” Central Asian Survey, in press, Tables 1 and 2; Etienne, , L'Afghanistan, p. 39. The latter is the source for the number sent abroad, which one reviewer felt was high. The author has not been able to find any corroborative data. It is not clear whether the number sent abroad includes those sent for military training.Google Scholar

¹³ Henry, S. Bradsher, Afghanistan and the Soviet Union (Durham, N.C., 1983), p. 29.Google ScholarHasan, Kakar, “The Fall of the Afghan Monarchy in 1973,” International Journal of Middle East Studies, 9 (1978), 212, gives higher figures of 7,000 trained in the USSR and Czechoslovakia, with 600 trained in the United States.Google Scholar

¹⁴ For a fuller treatment, see Rubin, “The Old Regime in Afghanistan.”

¹⁵ The main source was Ludwig, Adamec, A Biographical Dictionary of Afghanistan (Graz, 1987), supplemented by an interview with Hafizullah Karzai.Google Scholar

¹⁶ Khalq (the masses) and Parcham (the flag) were the names of the factions'newspapers. On factionalism in the PDPA. see Anthony, Arnold, Afghanistan's Two-Party Communism: Parcham and Khalq (Stanford, 1983);Google ScholarOlivier, Roy, “Le double code afghan: marxisme et tribalisme,” Révue française de science politique (12 1986), 846–61;Google ScholarRaja, Anwar, The Tragedy of Afghanistan: A First-hand Account (London, 1988). The latter source, based mainly on interviews carried out by the exiled Pakistani author while in prison with some Khalqi leaders, including the family of Hafizullah Amin, must be used with some care.Google Scholar

¹⁷ Published information on the PDPA leadership came from Adamec, Biographical Dictionary; Arnold, Afghanistan's Two-Party Communism; and press reports of appointments. I supplemented these with information taken from unpublished data bases compiled by Anthony Arnold and David Katz, which I was kindly permitted to consult, and several interviews with Miagol, the Kabul government's chargé d'affaires in Washington during 1989–90, who has been a member of PDPA-Parcham since 1967.

¹⁸ For a participant's memoirs of this period, see Mohammad, Eshaq, “Evolution of Islamic Movement in Afghanistan, Part 1: Islamists felt need for a party to defend Islam,” AFGHANews, 5 (01 1, 1989), 5 ff.Google Scholar

¹⁹ lbid., 8; Sayyed, Musa Tawana, “Glimpses into the Historical Background of the Islamic Movement in Afghanistan, Part (4),”AFGHANews, 5 (05 15, 1989), 5 ff.Google Scholar

²⁰ Ibid., 5.

²¹ I have also included some prominent individuals such as Maulana Faizani and Minhajuddin Gahiz who were part of a broader Islamic movement but did not join the formal organization. I have not included Sibghatullah Mujaddidi, of the famous family of Naqshbandi pirs. Mujaddidi, whose family represented the most Islamic wing of the old regime, was a prominent activist who was arrested by Daud in 1959. His brother Harun was active enough in the Ikhwan in Egypt to have been imprisoned by Nasser, and he was also connected to the Brotherhood himself. Mujaddidi has had good relations with Rabbani (with whom he shares a Naqshbandi connection) and bitterly antagonistic ones with Hikmatyar. In Pakistani exile, however, whenever the resistance parties split between Islamists and traditionalistnationalists, he always joined the latter. Hence I have not included him among the Islamic revolutionaries. His case illustrates, however, that the distinction between the Islamic revolutionaries and the Islamic establishment is not so absolute in Afghanistan as in some other Sunni countries.

²² Published sources included the memoirs of Sayyed Musa Tawana and Mohammad Eshaq (respectively the deputy leader and a political officer of Jamiat), both published in the Jamiat English language publication, AFGHANews; a series entitled “Who's Who in the Mujahideen”published in various issues of AFGHANews; Adamec, , Biographical Dictionary; Olivier Roy, Islam and Resistance in Afghanistan (Cambridge, Eng., 1986); Edwards, “The Evolution of Shiʿi Political Dissent in Afghanistan,” which also includes information on Sunni Islamists. I supplemented the published materials with information collected in interviews with Mohammad Eshaq, Naim Majrooh, Sultan Mahmud, and Abdul Jabbar Sabet.Google Scholar

²³ On the contrast of “mentalities” with ideologies, see Juan, J. Linz, “Totalitarian and Authoritarian Regimes,” in Fred, Greenstein and Nelson, Polsby, eds., The Handbook of Political Science vol. 3 (Reading, Mass., 1975), pp. 266–70.Google Scholar

²⁴ Pierre, Centlivres and Micheline, Centlivres-Demont, Et si on parlait de l'Afghanistan? (Neuchâtel and Paris, 1988), pp. 229–46.Google Scholar

²⁵ Published in Khalq on April 11, 1966. Reproduced in a translation obtained by the U.S. Embassy in Kabul in Anthony, Afghanistan's Two-Party Communism, pp. 137–48.Google Scholar

²⁶ Reproduced in ibid., pp. 149–59.

²⁷ According to Dr. Tawana, they chose the name Jamʿiyyat for the movement “because it resembled the word ‘Jamaat’ in the name of ‘Jamaat Ikhwan Muslemeen’ of Egypt and ‘Jamaat Islami’ of Pakistan but was also distinct from both” (Tawana, , “Glimpses, Part [4],” p. 5).Google Scholar

²⁸ Roy, , Islam and Resistance, p. 70.Google Scholar

²⁹ Fashradah-ʾi Hadaf va Marām-i Jamʿiyyati IslāmĪ-i Afghānistān (Summary of the Aims and Program of the Islamic Society of Afghanistan) (n.p. [Peshawar?], n.d. [1978 or 1979 by internal evidence]); Marām-i Ḥizb-i lslāmĪ-i Afghānistān (Program of the Islamic Party of Afghanistan) (n.p. [Peshawar?], 1365 A.H./1986–87). The latter is a fifth printing of a program that may have been written before 1978, as it mentions neither the 1978 coup nor the Soviet invasion.

³⁰ For this use of the term “jāhiliyya” see Marām-i, Jamʿiyyat, p. 8;Google ScholarMarām-i Ḥizb, p. 6, and elsewhere.Google Scholar

³¹ “Niẖām-i zindigĪ-i mardum bar pāyah-yi muʿtaqidāt-i ānhā istivār na-mibāshad.” (Marām-i Ḥizb, p. 5.)Google Scholar

³² Ibid., p. 38: “In niẖām fashradah-ʾi tamām-i mafāsid, maẖālim va biʿadālatihā-i niẖāmhā va sistimhā-yi ghayr-i Islāmi ast.”

³³ “Bidūn-i iṣlāḥ-i kāmil-i niẖām-i ḥhākim, nah iṣlāḥ-i fardi imkān pazĪr ast.”(Marām-i Ḥizb, p. 7.)Google Scholar

³⁴ Ibid., pp. 51, 69.

³⁵ Ibid., pp. 13, 59; the latter quote is: “PālĪsi-i iqtiṣādi-i kishvar bih ishtirāk-i namāyandigān-i muntakhab-i millat tarṭĪb gardidah, baʿd az taṣvib-i majlis-i shūrāʿamali migardad.”

³⁶ Roy, , Islam and Resistance, pp. 77–78.Google Scholar

³⁷ Hamied, N. Ansari, “The Islamic Militants in Egyptian Politics,” International Journal of Middle East Studies, 166 (1984), 140.Google Scholar

³⁸ Ibid.

³⁹ Asia Watch, Afghanistan: The Forgotten War: Human Rights Abuses and Violations of the Laws of War since the Soviet Withdrawal (Washington, D.C., 1991), pp. 99–124.Google Scholar

⁴⁰ “Ustad Umer Al-Talmesani,” The Mujahideen Monthly (05 1986), 15 ff.Google Scholar

⁴¹ Hikmatyar has taken this position verbally at times, but in view of the massive aid he has received from the United States, Afghans do not take his statements on this subject too seriously. The leader of Jamāʿat al-Daʿwā. Jamil al-Raḥman, was assassinated in Eastern Afghanistan by an Egyptian on August 30, 1991. This followed a battle between him and Ḥizb-i Islami for control of the Kunar Valley (AFGHANews7 [09 15, 1991], 7–8).Google Scholar

⁴² Marām-i Jamʿiyyat, pp. 10–11, says, “Jamiat has no inherent opposition to any individual or group” and states its willingness to cooperate with any that share its general goals. On p. 12 it calls for a union of “all believing and valiant compatriots, courageous youth … of ulama … of brave and valiant [military] officers … of those educated in modern sciences … of honorable and faithful peasants … of… workers … and of the whole believing nation” without insisting on a leading role for itself.Google Scholar

⁴³ Of course, the Pashtun predominance in the Parchami Politburo requires that one accept the Pashtun identification of Dr. Anahita Ratibzad and Babrak Karmal. Karmal and Ratibzad are both of ethnically mixed, Persianized, Kabuli backgrounds, and neither has any tribal ties.

⁴⁴ This is consistent with findings about the social background of “Islamic militants” and leftist activists in Egypt and Iran. See Saad, Eddin Ibrahim, “Anatomy of Egypt's Militant Islamic Groups: Methodological Note and Preliminary Findings,”International Journal of Middle East Studies, 12 (1980), 423–53;Google Scholar on Iran, Ibrahim Cites Ervand, Abrahamian, “The Guerilla Movements in Iran, 1963–1977,” MERIP Reports, 86 (03–04 1980), 3–15, especially Tables I and II, p. 5.Google Scholar

⁴⁵ The lone exception is the Khugiani tribe, from which Mawlawi Yunus Khalis comes. This tribe is Durrani by descent, but has been settled so long among the Ghilzai of Ningrahar that it no longer has any links to the main Durrani tribes. I have classified it here with non-Durrani tribes, as it lives outside the Durrani homeland.

⁴⁶ This analysis was carried out on all Central Committee members with adequate data (N = 99) with ethnicity coded as a dichotomy (Pashtun/non-Pashtun). The apparent relationship of ethnicity to factional membership is due to the fact that most (71 percent) of the Kabulis in the PDPA Central Committee were non-Pashtun, which is consistent with the population of Kabul, and most of the nonKabulis (80 percent) were Pashtun. Non-Pashtun educated provincial youth seem largely to have joined either the “ Maoist” organizations or the Islamists.

⁴⁷ The lone Khalqi from the elite schools was Muhammad Ismaʿil Danish, a Qizilbash Shiʿa from the Chindawul district of Kabul, who is thus triply (birthplace, ethnicity, education) unusual among Khalqis.

⁴⁸ The former Deputy Leader of Ḥizb, Qazi Muhammad Amin, who formed his own splinter party after a dispute with Hikmatyar, also attended a government madrasa.

⁴⁹ See Ibrahim, “Anatomy of Egypt's Militant Islamic Groups” and Abrahamian, “The Guerilla Movements in Iran.”

⁵⁰ See Table 10 in Rubin, “The Old Regime in Afghanistan.”

⁵¹ Le Monde (Paris), 05 7, 1991.Google Scholar

⁵² “Statement by the Secretary-General on Afghanistan,” United Nations, May 21, 1991.

⁵³ See Ashraf, Ghani's comments on “Panel E: Impact of Foreign Aid on Relations between State and Society,” WUFA (Writer's Union of Free Afghanistan), 5 (10–12 1990), 117–33.Google Scholar