3
agenda. It will, however, be appropriate to discuss what evidence was available to Smith to
confirm or falsify the theory, and to discuss the way he dealt with the available evidence.
The four stages in the Wealth of Nations and the Lectures on Jurisprudence
There is a clear, if implicit, reference to the four stages theory as early as the fourth paragraph
of the Wealth of Nations in Smith’s famous contrast between living standards in ‘savage
nations of hunters and fishers’ and ‘civilised and thriving nations’ (WN Intro.4), that is,
between the first and last of the four stages. That reference is typical of many others in the
Wealth of Nations in that the stages theory is clearly there in the background, but is not
spelled out in its own right. There is one substantial section of the book in which all four
stages are named, explicitly defined, and contrasted with each other (WN V.1.a and b), but
even there the stages are simply compared in a static fashion with no explicit claim that one
follows on from another, apart from a description of each stage as more ‘advanced’ than its
predecessor.
The main source for Smith’s use of the four stages theory, however, is the Lectures on
Jurisprudence, consisting of two sets of student notes on his lectures, relating to different
years. There are clearly difficulties in relying on the notes since we do not know how
complete or accurate the y are. The material was presented in a different order in the two
years, and one set of notes is longer, and therefore presumably fuller, than the other. There is,
however, reassuringly close agreement between the two note-takers in terms of substantive
content, suggesting that the notes give a good impression of the content of Smith’s lectures.
The fact that these are lectures, and lectures on jurisprudence, is a more substantial problem.
They are not a finished and considered work prepared for publication, and they are not a
work of history, still less a complete history of the world. Where there are gaps in coverage
or argument, it may simply be that Smith omitted things that were not directly relevant to the
course, or not suitable for that particular audience, or because of lack of time. The lectures,
however, together with the rather scattered material from the Wealth of Nations, are what
there is to go on.
3
Smith’s terminology and definitions can be confusing.
In the first stage, Smith’s ‘age of
hunters’, subsistence depended on the ‘wild fruits and wild animals which the country
afforded’ (LJ(A) i.27). People in this stage would now be called hunter-gatherers. The second
stage, the ‘age of shepherds’ or of ‘pasturage’ (LJ(B) 149), is characterised by the herding of
animals, but not the tilling of the soil. The animals herded by ‘shepherds’ need not be sheep.
What is critical in Smith’s account is that the domestication of animals came before the
domestication of plant s, to make a distinct stage in development. ‘We find accordingly that in
almost all countries the age of shepherds preceded that of agriculture’ (LJ(A) i.29). This
3
There is nothing directly relevant to the four stages theory in the Theory of Moral Sentiments.
4
stage, Smith thought, was typically nomadic – when the pasture in one area was exhausted,
shepherd and flock moved on. In Smith’s story the ‘age of agriculture’ or the ‘age of farming’
(LJ(B) 149) added (arable) farming (tillage, the cultivation of the soil), though the keeping of
animals and the eating of meat certainly continued. Agriculture in this sense required
investment in clearing and cultivating the land, and allowed food supply and population to
increase. I shall use the word ‘agriculture’, as Smith did, to mean settled agriculture with a
large arable component, and ‘pasturage’ to mean animal husbandry without tillage.
The commercial stage is different in that it is not defined by the main source of food.
Commerce (trade) plays some role in all stages of society, while the commercial stage, as
Smith defined it, is a development of the agricultural stage. A simple agricultural system
might have ‘little foreign commerce’ and only ‘coarse’ manufactures produced in the
household (WN V.i.a.6). A division of labour and corresponding pattern of trade develops bit
by bit (LJ(A) i.31), with no clear dividing line at which society becomes ‘commercial’.
Conjectural history
In his biographical account of Smith’s life, his friend Dugald Stewart described Smith’s
method as ‘conjectural history’, initiating a debate about the methodological basis of Smith’s
history which has continued ever since. Discussing Smith’s Dissertation on the Origins of
Language and, more generally,
his view of history, Stewart noted the lack of direct historical
evidence about (at least) the early stages of development from ‘rude tribes’ to contemporary
society, and remarked that:
In this want of direct evidence, we are under a necessity of supplying the place of fact by
conjecture; and when we are unable to ascertain how men have actually conducted
themselves upon particular occasions, of considering in what manner they are likely to
have proceeded, from the principles of their nature, and the circumstances of their external
situation.
…
To this species of philosophical investigation, which has no appropriated
name in our language, I shall take the liberty of giving the title of
Theoretical or
Conjectural History. (Stewart 1980 293)
This points to an important question: how much of Smith’s account of history is based on
evidence and how much is ‘conjecture’, invented to fill the gaps? Meek (1976 231) read
Stewart’s comments as ancestral to a line of criticism that accuses Smith of neglecting
historical facts in favour of theory. Thus, Coleman has remarked that ‘historical evidence was
of secondary importance in [Smith’s] grand design of a comprehensive system’ (1980 775),
and Wightman (1975 54) accepted a role for gap-filling in the absence of direct evidence, but
complained that Smith left the reader in doubt where fact ended and fiction began. Stewart
himself did not intend any such criticism – he saw conjecture as a means of filling the gaps,
and suggested that otherwise disconnected pieces of evidenc e might act as a check on the
story (1980 293). Indeed, one could reasonably argue that historical evidence never tells the