Agricultural Writing: 1523 to 1788

Between the 16th and 18th centuries, agriculture in England underwent numerous institutional and technological transformations (Allen Reference Allen1992). Crop yields significantly improved throughout the 16th century, leading to a 71% increase in population (Allen Reference Allen1999; Lowry Reference Lowry2003). Yields continued to improve in the 17th century, doubling between 1520 and 1651 (Allen Reference Allen1999; Wilson Reference Wilson1984). After enclosure began in earnest in 1700, farm sizes grew dramatically (Allen Reference Allen1988). Additionally, literacy rates among the farming classes increased over this period, creating a new audience for writing on agricultural topics (Lowry Reference Lowry2003). It is in this postmedieval period that weeds become part of the conversation in agricultural manuals. Agricultural writing has a long history in England dating back to the medieval period, but clear identification of plant species as weeds does not become part of the tradition until the 16th century, when Sir Anthony Fitzherbert (1470 to 1538), a judge in the English courts, succeeded his father as Lord of Norbury. As a wealthy country gentleman, he could afford to experiment with various agricultural practices without worrying about failure, leading to the 1523 publication of his agricultural treatise titled The Boke of Husbandry (Fitzherbert and Skeat Reference Fitzherbert and Skeat1882; see also Fussell Reference Fussell1947). While this text does not focus solely on arable agriculture, touching also on topics ranging from horse breeding to managing household accounts, it does contain a single chapter titled “To Knowe Dyuers Maner of Wedes” (To know diverse manner of weeds). In a short passage, Fitzherbert states that in late May and early June, one should prepare to “wede thy corne” and describes ten types of weeds found in the wheat fields, many of which can be identified from his short statements. For example, he offers the following description of corn cockle (Agrostemma githago L.): “Cockle has a long small leaf, and will bear five or six flowers of purple color, as broad as a groat, and the seed is round and black.” Although the English word cockle in this period can refer to either darnel (Lolium temulentum L.) or corn cockle, Fitzherbert’s description of the purple flowers clearly identifies this plant as the latter.

Other plants are more difficult to identify, as Fitzherbert uses more obscure names. For example, he lists a plant that seems to match the description of cornflower (Centaurea cyanus L.), as he describes it as having “a blue flower, and a few little leaves, and has five or six branches, flowered in the top,” but he names it as “hawdod,” a plant name found in only one other source in the early modern English corpus (Biggam Reference Biggam1994). Similarly, his description of wild oat (Avena fatua L.) uses the word “drake” to identify the plant, a word that in Middle English can refer broadly to weeds, to darnel, or to corn cockle, as well as to wild oat. The identification with wild oat is secure in this context due to the description of the plant (it “is like unto rye, until it begins to seed, and it has many seeds like fennel-seeds, and hangs downward, and it may well be suffered in bread, for there is much flour in the seed: and it is an opinion that it comes from rye”) and the identification of corn cockle and darnel separately within the larger passage.

Many subsequent early modern agricultural treatises identify specific plants as weeds without offering detailed descriptions of those plants, such as works by Gervase Markham (Reference Markham1614, Reference Markham1620), Samuel Hartlib (Reference Hartlib1651), Walter Blith (Reference Blith1653), John Smith (Reference Smith1673), Leonard Meager (Reference Meager1697), Timothy Nourse (Reference Nourse1700), John Laurence (Reference Laurence1727), and Richard Bradley (Reference Bradley1727). Markham (Reference Markham1614) is an exception in some cases, describing common knotgrass (Polygonum aviculare L.) as “a long running weede with little round smoth leaues, and the stalke very knotty and rough winding and wreathing one seame into another very confusedly, and groweth for the most part in very moist places,” but generally he offers only the names of plants and describes each as either a “weede” or an “hearbe.” There is an acknowledgment by some of these authors that different environments produce different weeds; Markham (Reference Markham1620, 7) states that ground that lacks grass but produces “Hemblocke, Docks, Mallowes, Nettles, Ketlocks, and such like” is good for cereals, as is ground that produces “Reede, Rushes, Clouer, Daysie, and such like,” while Smith (Reference Smith1673, 32) similarly states that some soils “naturally produce Weeds, as Mallows, Nettles, Docks, Hemlocks” and others grow “Daisies, Clover, Charnock, Mustardseed, Rushes, &c.” Overall, these works from the 17th century do not provide descriptions of weedy plants, nor do they provide any attempts at comprehensive listings of weed species and methods of control.

One text that does not quite fit into the pattern laid out by other 17th-century agricultural manuals is The Art of Simpling by William Coles (Reference Coles1657). More a botanical manual than an agricultural treatise, this book suggests a division between weeds (“Chickweed, horehound, Archangell, Cleavers, Grounsell, Nettles, Hemlock, Bindweed, Poppy … Cockle, Mayweed, etc.”), useful plants found in the fields (“Tormentill, Agrimony, Fluellin Woodsorrell, etc.”), and grasses (“Satyrions, Knapweed, Scabious, Yarrow, Pearl-grasse, Dog-grasse, Trefoile, Daysies, Crowfoot, Ladies Bedstraw, etc.”) (16). Plants found “amongst the corn” are listed as “Blewbottles, Poppies, Restharrow, Fumitory, Sheepherds needle, Mayweed, Cockle, Corne Marigold, Pimpernell, Cowparsnep, bindweed, Sow-Thistles, etc.” (18). None of these plants receives a detailed description, but the distinction drawn between weeds, useful plants, and grasses and the identification of plants found in arable fields indicate the application of the concept of the weed by early botanists and provide insight into what weedy plants may have been present in 17th-century fields.

In the mid-18th century, there was increased interest in identifying specific plants as weeds. William Ellis (Reference Ellis1741) included a chapter on “Weeds, and Their Cure” in volume 5 of his treatise The Modern Husbandman. This volume, which provides advice for farmers in the month of May, advertises two new types of plows for weed control (conveniently for sale through Ellis’s own business) and describes 29 types of weeds, including many (but not all) of those found in Fitzherbert’s text. (See Table 1 for a summary of weeds that appear in multiple manuals.) Unlike Fitzherbert, Ellis was a farmer rather than a country gentleman, and the more extensive list of weedy plants may reflect his familiarity with his fields (Thick Reference Thick2022). The 1758 edition of Thomas Hales’s The Compleat Body of Husbandry, which spans four volumes and more than a thousand pages, similarly contains a section in the third volume dedicated to weed control. The grouping of specific species itself is not intended to be comprehensive but instead offers a beautifully detailed page of black-and-white illustrations of common weed species categorized by annuals and perennials (485). In his 1759 continuation of his treatise, he goes on to outline the uses of various “unregarded” plants. Although he does not name these plants to be weeds as a group, certain examples, such as bastard alkanet [Buglossoides arvensis (L.) I.M. Johnst.], are named as such (56).

Table 1. Weeds mentioned in multiple published sources from the period 1523 to 1788.

A more botanically inclined text from the mid-18th century is another clear example of a forerunner to the modern weed identification manuals. Bradley’s (Reference Bradley1747) Dictionary of Plants gives detailed descriptions of several species identified in agricultural sources as weeds, although the dictionary itself is not consistent in the application of that term. For example, common groundsel (Senecio vulgaris L.) is identified as an “evil Weed” in the second volume of the dictionary, while other plants, such as corn marigold [Glebionis segetum (L.) Fourr.], which is identified as a weed in texts as early as Fitzherbert, in volume 1 is not given that label.

The Complete Farmer (Anonymous 1766) offers an even more detailed look at weeds and weed control in early modern England. This text defines a weed as “any plant growing in a field different from what the farmer intended” before detailing the ecological reasons weeds are so destructive; explaining annual, biennial, and perennial life cycles in weedy species; outlining methods for destroying weeds; listing common British weeds by where they are found (arable land vs. pasture) and when they are likely to be troublesome; and describing the physical appearance of each of the species mentioned. A slightly later attempt to comprehensively list specific species as a coherent group of “weeds” comes from 1788, when a Mr. William Marshall provided a written list of “Weeds and Vermin” in his account of the rural economy of Yorkshire. He focused on methods for the destruction of these plants rather than describing their morphology in great detail but generally offered enough information that someone familiar with weeds in arable fields would be able to determine which plants he is referencing.

Herbals and Botanical Illustration: Early Medieval to Early Modern

While not explicitly agricultural in nature, illustrated herbals represent a long tradition of scientific writing in premodern Europe broadly and England specifically. Herbals are medical treatises filled with plant-based remedies for various ailments and date back to European antiquity as a genre of medical and botanical writing. In England, manuscripts of herbals containing identifiable botanical illustrations date back to the early medieval period. The earliest illustrated herbal from England, the 9th-century Cotton MS Vitellius C. iii held by the British Library, contains the text of the Old English Herbarium, which is a collection of Old English translations of various herbal medical treatments that originated with Latin authors (Van Arsdall Reference Van Arsdall2023). Later illustrated herbals in Latin with Old English glosses and commentaries from medieval Britain include Bodley 130 and Ashmole 1431, both produced in the late 11th century and held by the Bodleian Library at Oxford (Collins Reference Collins2000). Figure 1 reproduces an illustration of an unidentified knapweed species (Centaurea spp.) similar to the cornflower described in the 16th century by Fitzherbert (Reference Fitzherbert1523). The introduction of printing and in particular the use of woodblock printing for images to Europe in the 15th and 16th centuries created an explosion of illustrated herbals, particularly in the Netherlands and eventually in England as well (Chen Reference Chen2020; Moran Reference Moran and Valleriani2017; Rosenberg Reference Rosenberg2023). Figure 2 reproduces an illustration from an entry in John Gerard’s (Reference Gerard1597) herbal that likely refers to greater knapweed (Centaurea scabiosa L.).

Figure 1. Illustration of “curmelle,” an unidentified species of Centaurea. Fol.1r from Oxford MS Bodley 130.

Figure 2. Illustration of “great Centorie” or “Centaurium magnum,” likely greater knapweed (Centaurea scabiosa L.), from Gerard’s Herball (1597).

Two important elements of these medical writings implicate them in the development of weed identification manuals. The first and most obvious connection is the use of early botanical illustration of specific and identifiable plant species, a form of scientific illustration that became both more technical over the course of centuries and more common with the advent of printed illustration in the early modern period. The second element of these texts is the description of the habitats of the plants included within the herbals. As far back as the Old English Herbarium, plants are identified by the locations in which they grow, such as fields, meadows, ditches, or other specific places in the landscape (see, e.g., the description of lady’s mantle [Alchemilla vulgaris L.] in Van Arsdall Reference Van Arsdall2023).

These two elements develop together in particular during the early modern period. Gerard’s (Reference Gerard1597) early printed herbal frequently describes plants such as corn marigold as growing “among corne” and contains a printed illustration of the plants in question. A later herbal by Elizabeth Blackwell (Reference Blackwell1737–1739) similarly describes and identifies plants that are found in fields or “among Corn,” such as restharrow (Ononis spinosa L.), and provides detailed illustrations of these plants. Unlike in the agricultural treatises, these plants are not explicitly identified as weeds, and the word “weed” appears only in the names of plants, for example, mayweed (Anthemis cotula L.) named in both Gerard (Reference Gerard1597) and Blackwell (Reference Blackwell1737) as Cotula foetida.

Weed Identification Manuals: 1792 to the Present

What might be termed the first true weed identification manual or textbook appeared in the early 1790s, bringing together these various threads touching on noncultivated plants in agricultural contexts. Thomas Martyn, a lecturer in botany at the University of Cambridge, published in four volumes the Flora rustica: Exhibiting Accurate Figures of Such Plants as Are Either Useful or Injurious in Husbandry, the first volume of which was published in 1792 and the last in 1794. This text, which includes detailed botanical illustrations of each plant alongside scientific descriptions and observations of the plants, combines harmful flora with cultivated and beneficial plants in a manner duplicated more than half a century later in the United States by William Darlington (Reference Darlington1847; Byrd et al. Reference Byrd, Russell and Broster2023). Figure 3 reproduces the illustration of cornflower from volume 4 of Martyn’s (Reference Martyn1792–1794) treatise. The plant itself is identified as a common weed in the accompanying text. This is not to say that Darlington was inspired by Martyn; there is no evidence that he was aware of the Flora rustica, which was never completed and had only a limited print run in the 1790s (Gorham Reference Gorham1830, 197). Darlington was, however, a prolific collector of botanical texts, including works by British writers, such as Smith’s English Botany (volume 1 of which is attributed to James Sowerby [Reference Sowerby1790], who created the illustrations, but the text is by Smith [Gorham Reference Gorham1830]), Loudon’s 1829 Encyclopedia of Plants, and Hooker’s 1830 British Flora (Darlington Reference Darlington1826, Reference Darlington1829, Reference Darlington1832).

Figure 3. Illustration of cornflower (Centaurea cyanus L.) from Martyn (Reference Martyn1794).

Cross-pollination between England and American agricultural writing can be traced back to at least the late 18th century, documented through correspondence from the time of the Revolutionary War onward in letters from James Warren to John Adams requesting the purchase of a copy of Arthur Young’s treatise Rural Economy (Loehr Reference Loehr1937). Sir John Sinclair, the Scottish agriculturalist, corresponded with George Washington on the subject of agricultural improvement and the potential creation of an American Board of Agriculture from the early 1790s onward (Sinclair Reference Sinclair1831). Arthur Young similarly corresponded with Washington on agricultural improvement throughout the late 1780s and the 1790s, with Washington raising general concerns about weeds in several letters (Washington Reference Washington1803, 11, 16, 102). Concerns about specific weeds appear in American agricultural writing dating back to at least 1782, in the text Letters from an American Farmer, where the fictional narrator, “James,” expresses particular concerns about wild cucumber [Echinocystis lobata (Michx.) Torr. & A. Gray] alongside hogweed (Heracleum spp.) and other species (St. John de Crèvecoeur and Stone Reference St. John de Crèvecoeur and Stone1981, 365).

Later 19th-century publications were not as ambitious as Martyn. William Pitt’s treatise on weeds, for example, took the form of a letter to the British Board of Agriculture and was printed in a collected volume in 1806. A later essay by Benjamin Holdich (Reference Holdich1826, 1) criticized Pitt’s letter, claiming that Pitt “knew little of agriculture” and that his letter had many omissions and lacked “any practical arrangement.” Holdich furthermore claims to be the first to publish such an essay on weeds and their destruction. While this is patently untrue, the short length of Holdich’s essay in comparison to earlier, more exhaustive texts provided a model for later weed identification manuals across North America in particular. These manuals, often focused on specific regions, such as individual states or provinces in the United States and Canada, were frequently short bulletins or essays that offered practical advice for farmers on identifying and destroying weeds in their fields (e.g., Anonymous 1894; Hillman Reference Hillman1893). Other texts resembled Martyn (Reference Martyn1792–1794) or Darlington (Reference Darlington1847) in their format but focused entirely on weeds (e.g., Michener Reference Michener1872). A further extension of this genre, usually authored by women, contained illustrations and descriptions of these plants and identified them as weeds while being classified as botany rather than weed science, such as Jane Loudon’s illustrated manuscript on British wildflowers from 1846 or Anne Pratt’s manual on noxious plants in the British landscape from 1866.

In the late 19th and early 20th centuries, weed identification manuals became more targeted for distribution to farmers, appearing in the form of pamphlets or short bulletins (e.g., Anonymous 1894). Some examples continued to produce colored plates to illustrate the species (e.g., Clark and Fletcher Reference Clark and Fletcher1906), while others included early examples of botanical photography to aid with identification (e.g., Hillman Reference Hillman1893). Since the late 19th century, weed identification manuals have proliferated to the point that it would be impossible to review all the manuals for North America. As these texts have proliferated, they have become more and more specialized. There are manuals focused on weed seeds (Delorit Reference Delorit1970; Martin and Barkley Reference Martin and Barkley1961), manuals focused on annual or perennial weeds (Anderson Reference Anderson1999; Cuthbertson and Hall Reference Cuthbertson and Hall1997), manuals for weeds in certain cropping systems or habitats (Bonanno and Everts Reference Bonanno and Everts2007; McCarty Reference McCarty2008), manuals for specific states or regions (Beal Reference Beal1915; Neal et al. Reference Neal, Uva, DiTomaso and DiTommaso2023) — the list continues. In the 21st century, weed identification tools have moved into the digital realm with websites like WeedImages.org and apps like ID Weeds (University of Missouri Extension, n.d.). The important characteristics of these manuals (and digital identification tools) are that they contain detailed descriptions and/or illustrations of the species in question, they often describe the habitats of said species, and they identify these plants as weeds. These characteristics did not emerge out of nowhere in the 19th century but have a long history rooted in premodern English writing on botany and agriculture.