Chapter 3: Kṛṣi-Parāśara, Rain Forecasting and the Agricultural Treatises¶
Sanskrit learning produced only a handful of works devoted to field agriculture. The best known is a short verse manual called the Kṛṣi-Parāśara, “Agriculture according to Parāśara”. It is often presented as an ancient compendium of farming science; the published text is shorter, later and more practical than that (Section 1.5). This chapter describes what the Kṛṣi-Parāśara contains, as far as the editions of Majumdar and Banerji (1960) and Sadhale (1999) allow. It then sets the text beside the older and fuller rain lore of Varāhamihira’s Bṛhatsaṃhitā, the rain records of the Arthaśāstra, and the weather sayings of the regional languages. A section asks what such forecasting could and could not achieve. The chapter ends with the other agricultural treatises of medieval India and a brief look at farming books written elsewhere in the pre-modern world.
These treatises have to be read against the society that produced them, which is described more fully in Chapter 18. By the time the Kṛṣi-Parāśara was written, land grants to brahmins, temples and monasteries had spread settled rice farming across eastern and southern India, and the grantees held rights over the produce and labour of the cultivators (Section 18.3). The texts assign agriculture to the vaiśya and discourage the higher varṇas from ploughing, while much of the actual work was done by śūdra peasants, by dependent labourers, many of them from castes later treated as untouchable, and by women (Section 18.2). The farming manuals speak for the landholder who supervised this work, not for those who did it, and the reader should keep that in mind throughout this chapter and the next.
3.1 The Text, Its Attribution and Its Date¶
3.1.1 Editions¶
The Kṛṣi-Parāśara first became widely available to scholars in the edition and English translation of Girija Prasanna Majumdar and Sures Chandra Banerji, published by the Asiatic Society in its Bibliotheca Indica series in 1960 (Majumdar and Banerji 1960). Majumdar, a botanist with a long interest in Sanskrit plant literature, had earlier translated the Upavanavinoda (see Chapter 4). In 1999 Nalini Sadhale published a further text and translation, with notes, as the second of the Agri-History Bulletins of the Asian Agri-History Foundation (Sadhale 1999). These two editions are the basis of everything said about the text in this chapter. Where they leave a point uncertain, it is left uncertain here.
The text is short: a few hundred verses of simple, mostly unadorned Sanskrit, with no commentary. It reads as a working manual for a landholding householder rather than as a learned treatise. Its language is sometimes irregular, and some of its technical words for tools and operations are rare. Both features suggest a regional text written for practical use.
3.1.2 Who was “Parāśara”?¶
The name Parāśara belongs, in the epics and Purāṇas, to a Vedic seer, the father of Vyāsa, to whom the Viṣṇu Purāṇa and several other works are traditionally ascribed. A Parāśara was also cited as an early authority on astral science and omens: Varāhamihira’s Bṛhatsaṃhitā and its tenth-century commentary by Utpala refer to older works on rain and portents ascribed to sages such as Garga and Parāśara (Bhat 1981). Placing a farming manual under the name of a famous seer was a way of giving it authority. It tells us nothing about the actual author, who is unknown. The attribution to Parāśara should be treated as traditional, in the same way as the attribution of the Arthaśāstra to Kauṭilya.
3.1.3 Date and place¶
Some modern accounts place the text in the Gupta period or even earlier, but there is no evidence for this. Majumdar and Banerji (1960) placed its composition between about 950 and 1100 CE; other scholars have argued for a later date. The text cannot be dated precisely, and any date before the late first millennium CE should be regarded as unsupported.
Its place of origin is more securely indicated. The agriculture it describes is centred on wet rice: nurseries, transplanting, standing water in the fields and its drainage. The sequence of operations fits the monsoon regime of the lower Gangetic plain, and part of its vocabulary is eastern Indian. For these reasons the text is generally assigned to Bengal or a neighbouring region of eastern India (Majumdar and Banerji 1960; Sadhale 1999). It belongs, then, to the same broad world as the Pāla and Sena kingdoms, whose copper-plate land grants record villages, fields, ponds and rice lands in some detail (see Chapter 18).
Scholarly Note: What the Kṛṣi-Parāśara does not contain
Much that circulates under the name of the Kṛṣi-Parāśara is not in the published text. It does not contain:
-
a recipe for kuṇapajala, the fermented liquid manure made from animal remains. That preparation belongs to the Vṛkṣāyurveda of Surapāla and to the Upavanavinoda (Chapter 4);
-
a classification of twelve soils, a “soil-testing section” using earthworms, or a description of petrichor;
-
pesticide formulas using tobacco or custard apple. Both plants came from the Americas and reached India only after 1500;
-
tables of cloud names, wind directions or animal signs with numbered verses;
-
a division of the year into kharīf, rabī and zāid seasons, which are Perso-Arabic revenue terms (Section 9.12), or recommendations of particular rice varieties for deep water or uplands.
On such attributions in general, see Section 1.5.
3.2 The Praise of Agriculture¶
The text opens, after the customary invocation, with a praise of agriculture. One of its verses begins kṛṣir dhanyā kṛṣir medhyā jantūnāṃ jīvanaṃ kṛṣiḥ, “agriculture is blessed, agriculture is pure, agriculture is the life of creatures” (Majumdar and Banerji 1960; Sadhale 1999). These verses praise agriculture as the means of life of all creatures and set it above other ways of earning a living. Even learned men, the text suggests, depend on food, and food depends on the farmer. The theme is old: the Ṛgvedic gambler is told to give up dice and plough his field, and the Tamil Tirukkuṟaḷ gives a whole chapter to the ploughman as the linchpin of the world (Sections 2.7 and 2.12).
The Kṛṣi-Parāśara addresses a householder who owns land and cattle and is expected to watch over the work himself. Its warm praise of farming stands in contrast to the normative literature, which was more ambivalent about the plough (Section 2.9).
3.3 What the Text Covers¶
The contents may be summarised under the headings that both editors recognise. The order below follows the logic of the farming year rather than the exact sequence of the verses.
3.3.1 Forecasting the year’s rain¶
Rain is the first concern of the text, and forecasting takes up a large part of it. The methods are those of the astral sciences (jyotiṣa) applied to the farmer’s needs.
The rulers of the year. Each year, in this system, has a planetary “king” and “minister”, fixed by an arithmetical rule applied to the number of the Śaka year; the minister is the fourth planet counted from the king. The character of the rains, and so of the harvest, is predicted from these rulers: some planets as king are said to bring abundant rain, others scanty rain or drought (Majumdar and Banerji 1960; Sadhale 1999). The same idea survives in the annual almanacs (pañcāṅga) of many Indian regions, which still announce the “king” and “minister” of the coming year.
The cloud of the year. The text names four kinds of cloud, āvarta, saṃvarta, puṣkara and droṇa, and gives a simple arithmetical rule, applied to the number of the year, for finding which of them governs a given year. Each cloud is associated with a different quality of rain, from patchy or scanty to abundant. These four are the classic cloud names of Sanskrit literature; they are not meteorological cloud types in the modern sense of forms such as cumulus or stratus. Kālidāsa’s Meghadūta, near its beginning, addresses the cloud-messenger as born in the famous lineage of the Puṣkaras and Āvartakas.
Observations in particular months. Beside these calendrical rules the text prescribes observations of wind, cloud, rain and sky on fixed days of particular months, from which the rain of the coming season is to be inferred. What is seen in the months before the rains is read as a sign of what will follow during the monsoon. The underlying idea, that the weather of one season is “carried” into another, is shared with the doctrine of the “pregnancy” of clouds in the Bṛhatsaṃhitā (Section 3.4).
The quantity of rain. The text also treats the amount of the year’s rain, reckoned in traditional measures, and how it is to be known. Here too it stands close to the older astral literature, in which rainfall was measured by collecting it in a vessel of standard size (Sections 3.4 and 3.5).
The rain lore of the Kṛṣi-Parāśara is less detailed than Varāhamihira’s, and much of it is clearly derived from the same tradition. Its distinctive contribution is to place these forecasts at the head of a manual of field operations. For its author, judging the year’s rain was the first thing a cultivator needed to do.
3.3.2 Cattle and the cattle shed¶
Draught cattle were the farmer’s capital, and the text gives them close attention (Sadhale 1999). It describes the cattle shed: where it should be built, the need to keep it clean and to remove dung and urine, and the protection of animals from damp and disease. It also gives rules for the care of the animals, and prescribes auspicious times and rites connected with them. The tone is that of a householder protecting valuable property. Livestock husbandry in general, including the veterinary literature, is treated in Chapter 13.
3.3.3 Manure¶
The text gives a seasonal routine for cow-dung manure: in Māgha the dung heap is to be dug out, then dried in the sun and broken up, laid in a pit in each field in Phālguna, and spread at sowing time (Majumdar and Banerji 1960; Sadhale 1999). The text warns that a crop grown without manure will bear no fruit. It does not offer a theory of soil fertility, and it does not describe the elaborate fermented preparations of the horticultural texts. Its manure is the ordinary farmyard manure of a mixed rice-and-cattle economy. Soil and manuring are discussed more fully in Chapter 7, and the later scientific study of Indian manuring in Section 10.7.
3.3.4 The plough¶
One of the most frequently cited passages of the Kṛṣi-Parāśara names the parts of the plough and gives their dimensions in traditional units such as the hasta (cubit), vitasti (span) and aṅgula (finger-breadth). Among the parts named are the īṣā, the long pole or beam joining plough to yoke, and the yuga, the yoke itself. Several of the other names are rare words whose exact reference is uncertain, and the two editors do not always agree on them; they are not listed here. The plough described is the light wooden ard drawn by a pair of oxen, with an iron share, that remained the standard implement of Indian farming until the twentieth century (Randhawa 1980). The passage is valuable to historians of technology precisely because it is so plain: it records a real tool, not an idealised one. The parts of such a plough are shown in Figure 7.1, and the history of the plough in India is traced in Chapters 2 and 7.
3.3.5 Seed¶
The text gives rules for collecting seed from the harvest, drying it and storing it so that it remains fit for sowing. Seed is to be kept dry and free from damage, and seed of poor quality will not give a good crop (Majumdar and Banerji 1960; Sadhale 1999). The text also prescribes rites and auspicious times for taking out the seed and beginning to sow. The wider history of seeds, landraces and their storage is treated in Chapter 9.
3.3.6 Sowing, transplanting, weeding and drainage¶
The heart of the practical section concerns rice. The text describes sowing, raising seedlings, and transplanting them into flooded fields during the rains. It insists on timely weeding, and it gives instructions for the management of water in the field: water is to be kept in the field while the crop needs it and drained out at the proper stage so that the grain can ripen (Majumdar and Banerji 1960; Sadhale 1999). This attention to letting water out of the field, as well as holding it in, is characteristic of a deltaic region where too much water was as great a danger as too little.
No yields, seed rates or spacings are given that could be converted into modern figures, and the text times its operations by lunar months and nakṣatras rather than by named crop seasons.
3.3.7 Harvest, threshing and storage¶
The text closes the agricultural cycle with the harvest, the preparation of the threshing floor, threshing, and the storage of grain. These operations, too, are linked to auspicious times and to rites performed on the threshing floor. The details are practical rather than technical, and the text says nothing about methods of preserving stored grain against insects. Farmers’ storage practices are described in Section 9.14.
3.3.8 Rites of the farming year¶
Throughout the text practical instruction is interwoven with ritual. The first ploughing of the season, the first sowing and the bringing in of the harvest are all accompanied by offerings and observances, and auspicious lunar days and nakṣatras are prescribed for each operation. The text also describes a festival associated with the month of Pauṣa, the Puṣyayātrā (Majumdar and Banerji 1960; Sadhale 1999). Rites of this kind are typical of agrarian societies, and similar harvest festivals are still celebrated across eastern India. Agricultural festivals more generally are discussed in Chapter 2.
These rites were not decoration. For the author they were as much part of farming as ploughing and weeding: success depended on correct timing, and correct timing was understood in astral terms. A modern reader may separate the observational from the ritual parts of the text. Its author did not.
Cross-Reference: The nakṣatra calendar
The Kṛṣi-Parāśara, like all Indian agricultural texts, times its operations by lunar months, lunar days (tithi) and the twenty-seven or twenty-eight lunar mansions (nakṣatra). The structure of this calendar, its relation to the solar year and the seasons, and the history of the pañcāṅga are treated in Volume 12, Khagola Jyotiṣa. Because the lunar months drift against the solar year and the nakṣatra positions have shifted with the precession of the equinoxes, a nakṣatra named in a medieval text cannot be turned into a fixed modern calendar date without calculation.
3.4 Rain Forecasting in the Bṛhatsaṃhitā¶
The fullest Sanskrit account of rain forecasting is not in a farming manual but in the Bṛhatsaṃhitā of Varāhamihira, a sixth-century astronomer-astrologer associated with Ujjain in the Avanti region. The Bṛhatsaṃhitā is a compendium of the “natural” branch of astral science (saṃhitā): omens, portents, the signs of good and bad harvests, gems, animals, architecture and the care of trees. Varāhamihira repeatedly cites earlier authorities, and his chapters on rain record a body of lore that was already old in his time. The standard English translation is that of Bhat (1981), whose chapter numbering is used here.
Eight consecutive chapters deal with forecasting rain (Table 3.1). Two other chapters touch farming directly: chapter 54, on finding groundwater (dakārgala), treated in Chapter 8, and chapter 55, on the care of trees (vṛkṣāyurveda), treated in Chapter 4.
| Ch. | Title | Subject |
|---|---|---|
| 21 | Garbhalakṣaṇa | Signs of the “conception” of the clouds in the cool season |
| 22 | Garbhadhāraṇa | The sustaining of the cloud-embryos and the signs that they will yield rain |
| 23 | Pravarṣaṇa | The first rains, and the measurement of rainfall |
| 24 | Rohiṇīyoga | Forecasts from observations when the moon is in conjunction with Rohiṇī |
| 25 | Svātiyoga | Forecasts from observations when the moon is in Svāti |
| 26 | Āṣāḍhīyoga | Forecasts from observations at the Āṣāḍha full moon |
| 27 | Vātacakra | Forecasts from the direction of the wind |
| 28 | Sadyovarṣaṇa | Signs of immediate rain |
3.4.1 The “pregnancy” of the clouds¶
The central idea of these chapters is that the rains of the monsoon are “conceived” months earlier, in the cool season, and “born” when their time comes. The conception season begins in the month of Mārgaśīrṣa (November–December). An “embryo” (garbha) formed when the moon stands in a particular nakṣatra is said to be delivered as rain about 195 days later, when the moon is again in that nakṣatra (Bhat 1981). Six and a half months after the cool season brings us to the beginning of the rains, so the scheme links each day of winter to a day of the coming monsoon.
The forecaster was to watch the sky on the days of conception. Favourable signs include gentle winds from the north, north-east or east, a clear sky with soft, glossy clouds, haloes around the sun and moon, and mild thunder and lightning. Unfavourable signs, such as meteors, dust storms, thunderbolts and planetary eclipses, are said to destroy the embryo. An embryo accompanied by all five signs of wind, rain, lightning, thunder and cloud is expected to bring abundant rain. If heavy rain falls during the period of pregnancy itself, the embryo is thought to be lost, “miscarried” before its time (Bhat 1981).
The scheme is elegant and internally consistent, and it reflects an awareness that the seasons are connected. It is also, in its working detail, a calendrical construction. There is no known physical mechanism by which the weather on a given winter day determines the rain on a particular day half a year later, and there is no established evidence that forecasts made by this method have useful skill.
3.4.2 Measuring the rain¶
The chapter on the first rains (pravarṣaṇa) contains one of the clearest pieces of quantitative method in the whole tradition. Rain is to be collected in a round vessel one hasta (cubit) across, and the amount is to be expressed by weight: fifty palas of collected water make one āḍhaka (Bhat 1981). A vessel of fixed aperture is the essential principle of every rain gauge. Varāhamihira uses the measurement to judge whether the first rain of the season is good or poor, as one input to his forecasts. It is not described as part of a continuous record kept over many years.
3.4.3 Conjunctions, wind and immediate signs¶
The next three chapters prescribe observations at particular moments of the lunar calendar: when the moon passes Rohiṇī, when it passes Svāti, and at the full moon of Āṣāḍha, just before the main rains. The appearance of the sky, the direction of the wind and the presence of clouds or rain at these moments are read as signs for the season. The chapter on the “wheel of winds” (vātacakra) forecasts the year’s rain from the direction of the wind on the Āṣāḍha full moon (Bhat 1981).
The last chapter of the group, on immediate rain (sadyovarṣaṇa), is the most down-to-earth. It lists signs that rain will come soon: the look of the sky at sunrise and sunset, the colour and movement of the clouds, and the behaviour of animals, such as ants carrying their eggs and frogs calling (Bhat 1981). These are the kinds of short-range signs that farmers everywhere have used, and they are closest to what modern observers would accept as genuine indicators of approaching rain.
3.5 The Arthaśāstra and the Recording of Rain¶
Rain measurement in India is attested still earlier in the Arthaśāstra, the treatise on statecraft ascribed to Kauṭilya. In its present form the text is composite and dates from roughly 50 BCE to 300 CE; the attribution to a minister of Candragupta Maurya around 300 BCE is traditional (Olivelle 2013). The Arthaśāstra is discussed as a whole in Volume 18, Rājanīti and Dharmaśāstra; only its rain lore concerns us here.
In Book 2 the royal storehouse is to be equipped with a rain gauge (varṣamāna), a vessel with a mouth one aratni (cubit) wide (2.5.7; (Kangle 1965; Olivelle 2013)). The chapter on the superintendent of agriculture (sītādhyakṣa, 2.24), who managed the crown’s own farmland, then states the normal rainfall of different regions in the volume measure droṇa. The arid jāṅgala country is given sixteen droṇas, the wet ānūpa country half as much again; Aśmaka thirteen and a half, Avanti twenty-three; for the western regions (Aparānta), the Himalayan regions and lands watered by channels the amount is said to be unlimited (Kangle 1965; Olivelle 2013). The ancient droṇa cannot be converted reliably into millimetres of rain, so these figures cannot be compared directly with modern data. What matters is the practice: rain was measured at state storehouses, and normal amounts for different regions were known well enough to be written down.
The same chapter connects the forecasting of rain with the positions and movements of Jupiter and Venus and with changes in the appearance of the sun. It advises the superintendent to plan sowing according to the expected rain, choosing crops that need more or less water (Kangle 1965; Olivelle 2013). Here administrative record-keeping and astral forecasting stand side by side, as they do in Varāhamihira. The rest of the chapter on the superintendent is summarised in Section 2.11.1.
3.6 Folk Forecasting: Khanā, Ghāgh and Bhaḍḍarī¶
Alongside the Sanskrit texts, every farming region of India had, and still has, a stock of rhymed sayings about weather, seasons and crops. They were transmitted orally, among cultivators rather than scholars, and they were written down only in the nineteenth and twentieth centuries by local scholars and colonial collectors. Because they were oral, they are hard to date, and many versions of each saying exist. The examples below are described rather than quoted. No reliable critical edition exists from which a fixed wording could be cited, and inventing “typical” sayings would be no better than inventing verses.
Khanā. In Bengal the best-known body of such sayings is the Khanār vacan, “the sayings of Khanā”. Legend makes Khanā a learned woman astrologer, sometimes connected with the household of Varāhamihira himself, whose tongue was cut out because her predictions surpassed those of the men around her. The sayings ascribed to her are short Bengali couplets on the timing of ploughing and sowing, the signs of rain and drought, the right season for particular crops, the care of cattle and the management of the household. As recorded, their language is that of early modern Bengali; their origins may be older, but they cannot be dated. Similar collections exist in neighbouring regions, such as the sayings attributed to Ḍāk in Assam and Bengal.
Ghāgh and Bhaḍḍarī. In the Hindi-speaking north, weather and farming sayings are attributed to Ghāgh and to Bhaḍḍarī. Ghāgh is a figure of legend placed by different traditions in different centuries and regions. Bhaḍḍarī (Bhaḍalī in Gujarati and Rajasthani usage) is likewise legendary and is associated especially with weather prognostication. The sayings attributed to both deal with the signs of rain in the sky and in animal behaviour, the qualities of good and bad bullocks, the right moment for sowing, and the conduct of a sensible farmer. They circulate in Bhojpuri, Awadhi, Braj and Rajasthani forms.
It is notable that in Bengal the emblematic authority on weather and farming is a woman. Whatever the historical reality behind the legend, it reflects the fact that agricultural knowledge was carried by women as well as by men. Women did much of the work of seed selection, storage, transplanting and post-harvest processing (see Chapters 9 and 14).
These sayings differ from the Sanskrit texts in voice rather than in substance. They mix astral rules (rain on a particular nakṣatra, the weekday on which a month begins) with local observations of wind, cloud, dew, the behaviour of birds and insects, and the flowering of plants. Their value lies less in any single prediction than in what they show of farmers’ attention to their environment and of the way knowledge was condensed into memorable form.
3.7 What Could Such Forecasting Do?¶
A fair evaluation of this material must distinguish several kinds of rule that are mixed together in the texts, because they differ greatly in what they could achieve.
Calendrical and astrological rules predict the year’s rain from the planet that rules the year, from the cloud assigned to the year by arithmetic, or from the position of the moon on certain days. These rules have no physical basis. The number of the year cannot affect the monsoon, and a forecast derived from it can be right only by chance.
Long-range rules link observations in one season to rain in another, as in the doctrine of the pregnancy of the clouds. They rest on a sound intuition: the seasons are connected, and the state of the atmosphere in one season does influence the next. But the specific links proposed, from a particular winter day to a particular monsoon day, have no known mechanism. Modern long-range forecasting depends on large-scale patterns of ocean temperature and atmospheric pressure across the whole tropics, which no ground-level observer could have detected.
Short-range signs, such as rising humidity, the build-up of clouds, shifts of wind, haloes and the behaviour of insects, frogs and birds, can carry real information about weather in the next few hours or days. Farmers who read them well had a practical advantage. Such signs are local and short-lived, however. They can tell a farmer that rain is coming this afternoon, but not whether the monsoon will fail next year.
Measurement stands apart. The rain gauges of the Arthaśāstra and the Bṛhatsaṃhitā are genuine empirical instruments, and the regional figures of the Arthaśāstra show that measurements were used to describe normal conditions. There is no evidence, however, that continuous records were kept and analysed over long periods to test the forecasting rules.
The monsoon’s variability was a matter of life and death, and the failure of the rains brought repeated famines. Systematic seasonal forecasting began only in the 1880s, when H. F. Blanford of the newly founded India Meteorological Department issued forecasts based partly on Himalayan snowfall. In the early twentieth century his successor Gilbert Walker searched worldwide weather records for statistical predictors of the monsoon and in the process identified the Southern Oscillation, later understood as part of the El Niño system. The colonial famines of the late nineteenth century, when the rains failed and grain continued to be exported, showed that forecasting alone could not prevent catastrophe (Davis 2001). Even today seasonal monsoon forecasts have limited skill.
Debate: Did Indian rain lore anticipate monsoon science?
For: the texts treat rain as part of a seasonal system, measure rainfall with standard vessels, and record observations of wind and cloud; some short-range signs they list are physically meaningful.
Against: the core predictive rules are astrological; the long-range links have no mechanism and no demonstrated skill; there is no evidence of the systematic testing of predictions against records that defines a forecasting science.
Assessment: the texts deserve to be read as a serious attempt to bring order to the most important uncertainty in farmers’ lives. They are part of the history of meteorology, as similar weather lore is in Europe and China. They did not anticipate the physical understanding of the monsoon, and nothing is gained by claiming that they did.
3.8 Other Agricultural Treatises¶
The Kṛṣi-Parāśara is not alone. A small number of other works, in Sanskrit and in the regional languages, deal wholly or partly with agriculture and plant care; their dates, as far as they can be established, are given in Table 1.2 (Chapter 1). Those concerned mainly with trees and gardens are discussed further in Chapter 4.
3.8.1 The Kāśyapīya Kṛṣisūkti¶
The Kāśyapīya Kṛṣisūkti, “Kāśyapa’s verses on agriculture”, is attributed to the sage Kāśyapa. Like the Kṛṣi-Parāśara it is a Sanskrit verse text, but it comes from the south of India. Its date is uncertain; its editor proposes the later first millennium CE, but the evidence is slight (Ayachit 2002). The text gives particular attention to rice, the dominant crop of the river deltas and tank-irrigated lands of the peninsula, and to the management of water. It places agriculture within a framework of royal responsibility and religious merit, in which the ruler is expected to promote cultivation. Ayachit’s translation made the text accessible, but it has received far less scholarly study than the Kṛṣi-Parāśara. Its relation to the irrigation history of south India (see Chapter 8) remains to be worked out.
3.8.2 The Lokopakāra¶
The Lokopakāra (“For the benefit of the people”) of Cāvuṇḍarāya, written in Kannada around 1025 CE, is an early encyclopaedic work in Kannada. It brings together practical knowledge on topics such as astrology and omens, house-building, water-finding, cookery, perfumes, remedies and the care of plants, and an English translation was published by the Asian Agri-History Foundation (Ayangarya 2006). Its plant-care section belongs to the vṛkṣāyurveda tradition and is described in Section 4.1.5. The Lokopakāra matters for the history of agricultural knowledge because it shows Sanskrit technical material being recast in a regional language for a lay readership.
3.8.3 The Mānasollāsa¶
The Mānasollāsa, also called Abhilaṣitārthacintāmaṇi, was composed in 1131 CE at the court of the Western Cālukya king Someśvara III. It is an encyclopaedia of royal life in five sections of twenty chapters each, covering the acquisition and protection of the kingdom and the king’s enjoyments and pastimes (Shrigondekar 1925). Several parts concern plants and animals: the laying out of gardens, with plant-care instructions of the vṛkṣāyurveda type; horses and elephants; dogs and hunting; birds; fishing; and a celebrated section on cookery. The text describes the agriculture and husbandry of the court rather than of the village, but it shows how much practical knowledge of plants and animals was expected of a cultivated king. Its material on animals is used in Chapters 15 and 17, and its cookery in Chapter 14.
3.8.4 Other sources¶
Agricultural material is also scattered through the Purāṇas, notably the Agni Purāṇa, which has chapters on the care of trees and of animals; through the dharmaśāstra literature on land, tillage and its rules; and through inscriptions recording irrigation works and land grants. From the thirteenth century onward, Persian sources from the Sultanate and Mughal courts add much on crops, revenue and gardens; they are discussed in Chapter 18. Agricultural verse also exists in the regional languages, though it has received little critical study. In Tamil, the Ēr eḻupatu (“Seventy on the plough”), traditionally ascribed to the poet Kampaṉ, praises the plough and the cultivating Veḷḷāḷar; in Malayalam, the Kṛṣigīta, a verse work of uncertain, probably early modern date, presents the rice farming of Kerala as teaching given by Paraśurāma. Both are works of praise and of caste and regional identity as much as manuals, and their dates and attributions are traditional. For most of Indian history, however, agricultural knowledge was transmitted by practice and by word of mouth rather than in writing. The treatises are a small and socially narrow window onto it.
3.9 Farming Books Elsewhere¶
It can be helpful to set the Indian treatises beside agricultural writing from other pre-modern societies, provided the comparison is descriptive rather than competitive.
In the Roman world, Cato, Varro and above all Columella, whose De re rustica in twelve books was written in the first century CE, wrote for owners of estates worked largely by enslaved labourers. Columella covers soils, vines, olives, livestock, bees and gardens, and the management of the estate’s workforce. In China, Jia Sixie’s Qimin yaoshu (“Essential techniques for the common people”), compiled in the sixth century, is a long and systematic manual of field crops, vegetables, trees, animal husbandry and food processing, including fermentation; it quotes many earlier works. In al-Andalus, Ibn al-ʿAwwām of Seville compiled his Kitāb al-filāḥa (“Book of agriculture”) in the twelfth century, drawing on earlier Arabic, Greek and Andalusi sources; it is especially rich on irrigated gardens and fruit trees. The Byzantine Geoponika, compiled in the tenth century, is an anthology of older Greek agricultural writing.
Beside these, the Indian works have a distinctive profile. The Kṛṣi-Parāśara is very short and addressed to a single householder. The Indian texts are composed in verse for memorisation, closely tied to astral science and ritual, and concerned with rice and the monsoon rather than with the vine and the olive. The tree and garden texts discussed in the next chapter are, by contrast, closely modelled on medicine. Each tradition reflects its own ecology, economy and learned culture. None of these works records the whole agricultural knowledge of its society, and the cultivators whose practice they describe, enslaved estate workers in Italy, peasants in China and India, mostly left no writings of their own.
Summary¶
The Kṛṣi-Parāśara is a short Sanskrit verse manual, attributed by tradition to the sage Parāśara, probably composed in Bengal or eastern India around 950–1100 CE or somewhat later. After praising agriculture, it forecasts the year’s rain and then gives a householder’s rules for cattle, manure, the plough, seed, the cultivation of rice, the harvest and the rites of the year. Its rain lore draws on the older tradition recorded in Varāhamihira’s Bṛhatsaṃhitā and the Arthaśāstra, which also measured rainfall; the oral sayings of Khanā, Ghāgh and Bhaḍḍarī carry the same mixture of astral rule and local observation into the vernaculars. Measurement and short-range signs had real value; the astrological and long-range rules did not. The Kāśyapīya Kṛṣisūkti, the Lokopakāra and the Mānasollāsa show that agricultural writing in India was limited in volume but varied in region, language and audience.
Further Reading¶
The text and translation of Majumdar and Banerji (1960) remains the basic edition of the Kṛṣi-Parāśara; Sadhale (1999) offers a convenient translation with notes. Varāhamihira’s rain chapters can be read in Bhat (1981), and the Arthaśāstra’s chapter on the superintendent of agriculture in Kangle (1965) and Olivelle (2013). Ayachit (2002) translates the Kāśyapīya Kṛṣisūkti and Ayangarya (2006) the Lokopakāra. Gopal and Srivastava (2008) sets these texts in the wider history of early Indian agriculture, and Randhawa (1980) provides a broad survey.