A Taureg nomad in the Sahara.
Photograph by Carsten Peter, National Geographic. ©
"The nomadic strategy is one means by which people adapt to thinly spread resources and to the variability of the resources across space and over time. It is also a strategy for avoiding other deleterious environmental conditions, such as extreme heat or cold, disease, or predators. Furthermore, as human predators are always a risk, every adaptation is political, relating populations through power. Above all, the nomadic strategy is a means of maximizing, given circumstances, culturally defined objects, such as production, survival, and independence."
Nomads are hard to wrap your head around.--Philip Carl Salzman. Pastoralists: Equality, Hierarchy, and the State. (Boulder, CO: Westview Press, 2004). p. 39.
I have suggested before that we moderns face a formidable conceptual block when we try to understand the politics of ancient, classical, and medieval societies like the Roman Empire or the Han Dynasty. The central problem: our world is not like theirs. The Growth Revolution transformed the nature of civilization and the dynamics that drive it. If we do not understand the rules their world operated on then we are forever doomed to misunderstand what happened there.
The divide between the ever growing, post-industrial, mass societies of today and the static, agrarian, imperial centers that dominated the ancient world is vast. The gap between modern society and the nomadic pastoralists who terrorized these agrarian centers is even greater. Making sense of a particular nomadic empire or its leaders was a difficult task for the sedentary historians and statesmen of their own day; moderns face just as difficult a challenge. Fortunately, it is not a problem without solution. Just as the decisions of ancient statesmen are better analyzed once we understand the way agrarian empires functioned, so will the actions of ancient khagans begin to make sense only after we understand the inner workings of the nomadic societies they ruled.
Amongst the narrow band of scholars who concern themselves with such questions this is an intensely controversial topic. I shall lay the most controversial debates aside for a different day; I here I will focus on generalities most scholars agree upon before moving to more troubled waters.
Enter Philip Carl Salzman's Pastoralists: Equality, Hierarchy, and the State.
Dr. Salzman is the rare anthropologist who has no patience for the post modernist "intellectual fashion" that has degraded cultural anthropology to his current sad state.  His work has a strong empirical character to it; in many ways Pastoralists is a summary of life-time of fieldwork among and study of pastoralist peoples. Salzman began this fieldwork in Iran, and it is to the nomadic peoples of Iran he turns to introduce '20 observations' that inform the rest of this work. These observations were drawn from studying four Iranian nomadic groups - the Baluch of southeastern Iran, the Komachi of Kerman, in southern Iran, the Basseri of Fars, and the Yomut Turkmen living near Iran's northeastern borders -- but Salzman makes it clear that these observations are applicable to pastoral nomadic societies in general.
Salzman's General Observations on Nomadic Pastoralists 
1. Nomadism can be used to gain access to resources that are sparse.
2. Nomadism can be used as an opportunistic response to temporary availability of irregular and unpredictable resources.
Example: In Baluchistan both pasturage and water are extremely limited. If pastoralists were limited to one location their flocks and herds would soon use all resources available there. Mobility allows access to the additional resources the Baluch need to survive and thrive.
3. Nomadism is unlikely to be oriented to just one productive activity (i.e., pastoralism), but to several.
Example: Baluchi nomads who move from one place to another do not just keep flocks. They also tend to date orchards. These orchards can be left alone for large parts of the year and returned to at harvest time. Thus "nomadic mobility is not infrequently from a location of one productive activity, like pastoralism, to another" , such as aboriculture in Iran or trading along the old Sino-nomadic frontier.
4. Nomads do not "wander" in the sense of purposeless or directionless movement.
5. Nomads do not "wander" in the sense of continually moving to new lands.
Meaning: Pastoral nomadic movements are both purposeful and limited to certain areas and spaces. Unless threatened by disaster, nomadic pastoralists stick to a specific route or territory.
6. Nomadic migration patterns are more regular and repeated where macroeconomic features (e.g. seasons) determine availability of resource.
Meaning: Groups like the Baloch, who live in an environment where resources are sparse and whose distribution differs year to year, have less predictable migration routes than groups like the Komachi, where resources are more predictable. If you cannot count on a watering whole having the same amount of water every year, odds are you will not visit annually!
7. Nomadism is not tied in a determinate fashion with political structure.
8. Nomadism is not tied in a determinate fashion with economic orientation (e.g. specialization vs. self sufficiency).
9. Nomadism is not tied in a determinate fashion with land tenure or ownership.
Meaning: All three points meant to discourage people from concluding that nomads always choose nomadic pastoralism (or choose to abandon it) 'because' of one thing. There is no universal explanation for why people are nomadic or sedentary; common explanations for these choices, be they related to ruling political regimes, economic specialization, or land tenure have too many exceptions to be useful.
10. Nomadism is found both in isolated, remote regions and crowded, developed ones.
Examples: When we think of nomadic pastoralists, isolated peoples like the Baluch are usually what come to mind first. But other groups, like the Basseri, migrated through populated agricultural zones over the course of their travels. In the case of the Basseri, major cities were a part of the migration cycle because they offered the Bessari a chance to profit from selling their pastoral goods to settled populations.
11. Independent or qausi-independent nomadic tribes vary considerably in political structure.
Examples: The four nomadic peoples surveyed differ drastically in their political structures. The Bessari have strong, hierarchical chieftains who control migration routes of the entire people, while the Baluch political order is decentralized in the extreme, verging on anarchy.
12. Chiefships arise in nomadic tribes in confrontation with powerful external populations.
13. Nomadic pastoralism is politically centrifugal, militating against central and hierarchical power.
Note: These two points are related. (They also are a major theme of this book). The gist is this: states are not a natural feature of the nomad's tribal world. Collectivization is not necessary for economic reasons; because property is not permanent (land is not owned and surpluses in the only property pastoralism produces--animals--rarely last more than one generation) the nomadic world is an egalitarian one. The only thing that sets one nomad apart from another is the esteem of his peers. No one nomad has the coercive power to establish an autocracy - and they never will, so long as they are left to squabble amongst themselves.
External powers interfere with this dynamic. On the one hand, external powers usually demand to work with one person who they can recognize as the leader of the group. This recognition (often coupled with gifts, knowledge of outside affairs, or promises of assistance) elevates him above the other nomads. On the other hand, if the outside power is hostile to the group, then the nomads have reason to unite and cede power to whatever man they think can lead them to victory.
14. Nomadism is not determinatively tied to a particular kind of physiobiotic environment.
15. In some cases the political use of nomadism predominates over direct productive use.
Example: The Yomut Turkmen live in an area of rich abundance. Agricultural life is quite possible in the valleys in which they live. Many Persian farming vllages dot their traditional range, and in normal years the Turkmen did not migrate far (usually close to 25 miles during the winter season). The Yomut Turkmen are not nomadic because the environment forces them to move around, nor do they choose nomadism because it maximizes production. They remain mobile for a different reason: independence. When the Turkmen felt threatened by the Persian state or Persian villagers they would simply pack up and leave. Their mobility gave them a powerful political advantage over their neighbors.
16. Nomadism and pastoralism are capabilities that can and are taken up and set aside, sometimes temporarily and sometimes indefinitely.
17. Decreasing nomadic mobility is often a voluntary choice of nomadic peoples facing changing traditions.
18. Some nomads who might settle if they had access to good agricultural land continue thir nomadism because they have no viable alternative.
19. Sedentarization is not always a collective event.
20. Governments often want to settle nomads as a way of gaining greater control over them.
These last items are rather self-explanatory and do not need any additional explanation on my part.
Reflecting upon these observations, Dr. Salzman concludes:
"What kind of people, we may ask, are nomads? But this is the wrong question and it leads to false conceptions. In fact, nomads are not a kind of people but different kinds of people who use a particular strategy--that is mobility of household---in carrying out regular productive activities and defending themselves. We may better understand the lives of these people if we ask what they are trying to accomplish through this strategy, how they implement this strategy, why they do not choose apparent alternatives, and in what ways this strategy is tied to the environmental conditions in which they live. Nomads do not live to migrate; they migrate to live." (emphasis added). 
Salzman reached this conclusion by studying the nomadic peoples of the present, but the questions he raise are just as applicable to nomadic societies of the past. It is very easy to take nomadism of these peoples for granted. We must avoid this temptation. The first step in understanding the Xiongnu, Seljuqs, or Mongols is to ask why these people chose the nomadic strategy in the first place.
 Philip Carl Salzman. Pastoralists: Equality, Hierarchy, and the State. . (Boulder, CO: Westview Press, 2004). p. 154
 Every point in this list is taken from Chapter 2, "Agency and Adaptation: Pastoralists of Iran" (Salzman, Pastoralists, p. 17-41). I have paraphrased most, but not all, of these observations to make them more concise.
 Ibid., p. 24.
 Ibid., p. 40