The Evolution of Useful Things: How Everyday Artifacts-From Forks and Pins to Paper Clips and Zippers-Came to Be as They Are.

by Henry Petroski

  Perhaps because the use of saws requires more effort than many other tools, there is an especially great proliferation of styles. Both the felling and the pit saw evolved into excellent tools for their tasks, but they were certainly too large, heavy, and unwieldy to use in the shop of a cabinetmaker or a wheelwright. Thus smaller hand saws developed independently. Some of these were for cutting large planks and panels, but their blades proved still too thick and long to provide the accuracy desired in cutting close-fitting joints and the like. Thus, the back saw evolved with a reinforcing top edge to keep its thin blade from distorting while doing fine work. But none of these saws was suited for cutting curves, and so the coping saw evolved, with its very narrow blade fitted into a frame that does not get in the way. (Because the coping saw’s blade is necessarily thin and thus prone to buckling if pushed hard against the work, the blade cuts on the pull stroke like an Oriental saw.) But even this saw fails to work when very tight curves must be cut well inboard from the edge of a board or a panel. Thus, the fret saw developed, with an especially deep bow to its frame and a blade that can cut curves without being angled.

  These are just some examples of specialized saws, of course, but they serve to illustrate the evolution within the genre in response to the failure of existing saws to perform a task that arises naturally in the course of timber cutting and woodworking. Failure-driven evolution does not guarantee that each new form will be a success, however, and there are plenty of examples, in old toolboxes and patent files, of failed ideas. One of the problems with the growth of specialized saws was the tendency of users to acquire many separate implements, thus creating problems of capital investment and storage. It must certainly have been the perceived disadvantage of needing separate crosscut- and ripsaws that prompted one inventor to devise a single saw that had crosscut teeth on one side and rip teeth on the other side of its single and otherwise symmetrical steel blade. The handle on this duplex saw was naturally mounted symmetrically on the blade, so that it might serve equally for either set of teeth. Unfortunately, the saw did not work very efficiently, for the highly specialized handle on a traditional saw is set well on the back of the blade and angled for balance and for directing the thrust of the hand most effectively to the cutting edge. By ignoring the details of the highly evolved saw handle, the inventor virtually condemned his duplex model to be a functional failure on all counts and thus an evolutionary wrong turn into a dead end.

  One of the most often cited tools in the context of studies of the evolution of form is the ax, and David Pye has cited it as a prime example of form not following function, because to his thinking it may be ideal for producing wood chips but it is terribly inefficient for cutting down trees. Nevertheless, the ax is a tool that cries out to be considered in any theory of the design and evolution of artifacts. Originating in the Stone Age, it developed in effectiveness and form as new materials and new methods of hafting became available. By colonial times, the modern European ax was pretty well established and tradition-bound, and such extratechnological cultural inertia can fix the form of an artifact in its home territory in spite of its inefficiency and functional shortcomings. After all, there is no technological imperative in efficiency, which in any event is in the eye and hand of the tool beholder.

  The Europeans had long known that iron is not easily hardened to give a keen edge, and so eventually it became customary to weld a strip of steel onto the business end of the ax head so that the blade could be honed to a fine cutting edge that did not wear out so quickly as iron. But even with a sharp edge, the European ax would tend to twist if not held firmly when being swung, because the bulk of the metal was on the front side of the handle and thus was easily turned aside by any but the most direct and controlled of blows. In young America, however, with its abundance of forests and its pressing need to build houses and other wooden structures on cleared land, the inefficiency and touchiness of the European ax were not accepted so uncritically. As late as 1700, American-made axes still looked very much like their European ancestors, except that they began to incorporate a poll—the blunt end of an ax head that projects out from the back side. The addition of this feature not only made the ax head heavier, and thus capable of packing a bigger wallop, but also moved the center of gravity and percussion back from the blade and closer to the handle, and thus had a stabilizing effect on the ax head during the blow.

  By the end of the eighteenth century, the American polled ax was well developed, and it even included a longer cutting edge than its European counterpart. However, no artifact can ever be considered absolutely perfected for all the tasks in which it comes to be employed; an ax design that seemed to be all that one woodsman might want left something to be desired for another. For example, sharpening a dulled ax might not be so inconvenient for the farmer working near his barn or toolshed, where a grindstone was located, but it could be very inconvenient for the woodsman working far away from such a stone. The two-edged ax could thus enable the woodsman to range farther from his home base, because he would have to return only half as often to sharpen his tool.

  Many local variations in the ax head developed, perhaps because trees are different in different parts of the country. But there are always subjective aspects to using tools and other artifacts, and, whether by tradition, habit, or feel, each woodsman might be expected to find fault with different and unfamiliar axes and to find one ax to be an improvement over another. In the nineteenth century, there came to be a proliferation of ax styles in America, each varying slightly from the others in its form. The different styles were frequently distinguished by regional names, and may have been selected by users more for chauvinistic than for technological reasons. George Basalla has suggested the range of such localization of form by noting that in 1863 one manufacturer listed felling axes in varieties designated “Kentucky, Ohio, Yankee, Maine, Michigan, Jersey, Georgia, North Carolina, Turpentine, Spanish, Double-bitted, Fire Engine, and Boy’s-handled.” Within two decades, such a list could exceed a hundred items, each one implying some shortcoming of the others. The name “Kentucky ax,” for example, suggests the failure of others to deal with the special problems of the Kentucky woodsman.

  Hammers have also been the frequent focus in studies of technological form, perhaps in part because of Marx’s astonishment that Birmingham produced so many different kinds. The specialization of hammers is perhaps not so surprising when we recognize that, like saws, these tools get a tremendous amount of repetitive use by craftsmen, who thus have considerable thinking time on their hands. Whereas the vast majority of hammer wielders will tend to adapt to and accept the faults of their tools, the rare creative individual will spend time thinking about how to avoid a particular problem that he keeps encountering in the course of using a less-than-perfect tool for a particular purpose. (The observant and inventive toolmaker might also come up with improvements in his products. These even the lumbering craftsman would purchase immediately if he recognized the innovative features as removing a familiar, if unarticulated, shortcoming of an existing tool.)

  In even the most functional of artifacts, imagination and whimsy can affect form. A design patent was issued in 1898 for this anthropomorphic hammer. Another common tool, a pair of dividers, has frequently been formed with human legs—sometimes very shapely or very muscular ones—by which distances could literally be stepped off at the workbench. (photo credit 7.3)

  The astounding number of times a day that a hammer wielder can pound is familiar to anyone who has lived under, or even just in the same neighborhood as, a roof being reshingled. When a roofer once left his hammer behind on my new roof, I was struck by how worn the head was and how polished the wooden handle, at least where it was not taped to contain a crack.

  TYPES OF NAILS AND SPIKES. (photo credit 7.4)

  In the late 1800s, Benjamin Butterworth, nineteenth U.S. Commissioner of Patents, arranged and compiled illustrations from related patents to trace the evolution of artifacts. This page from The Growth o
f Industrial Art shows a great variety of nails, demonstrating the specialized forms of artifacts that can develop out of the failure of existing ones to perform a function as effectively as inventors can envision.

  A “collector’s handbook” entitled The Hammer: The King of Tools contains over a hundred pages of photographs, typically showing ten or twelve different styles per page, of odd and unusual hammers and hammerheads. Another two-hundred-page section of the book reproduces the drawings (four to a page) of U.S. patents granted between 1845 and 1983 for improvements and variations on hammers and hammerlike tools, each one capable of performing some function all other hammers fail to—at least in the mind of the inventor in question. To be sure, many of the variations among tools are covered by design patents, which are granted for novel appearance alone. Still, even though no functional improvement need be present for a design patent, the most successful applications are for artifacts that show “a startling or unique new appearance.”

  But, for all the personalization and specialization of hammers over the years, late-Victorian merchants were apparently not averse to selling a given style to as wide a range of users as possible. In its 1895 catalogue, Montgomery Ward offered hammers to a strange list of bedfellows indeed, and one light-headed hammer with a large angled claw was clearly to be used in place of other models that defaced baseboards and were incapable of removing stubborn tacks. This was, according to the telegraphic catalogue description,

  The best tack hammer on the market. Will drive tacks without defacing base boards, and will extract the most obstinate tacks without effort. The favorite hammer with upholsterers, carriage trimmers, bill posters, carpet layers, undertakers, photographers, dentists, picture frame makers and cigar dealers. For use about the house it is unsurpassed.

  This is a versatile hammer indeed, and one wonders how many hammer styles Montgomery Ward, let alone all of Birmingham, might have had to make, had undertakers and dentists, for example, not been able to work with the same model. That a single hammer could have so many divergent specialized users does suggest that there is a limit to diversity, and that the limit represents a balance rather than a conflict between utilitarian and economic means and ends.

  8

  Patterns of Proliferation

  Among the most talked-about items at antique shows are the odd and unusual pieces of old silverware whose handles clearly indicate they fit a familiar place-setting pattern but whose intended use can be a matter of considerable speculation. Dealers and collectors alike debate function more stubbornly than value, purpose less convincingly than price. The uninitiated does not have to eavesdrop long to experience utter confusion over whether one fine-looking piece is for serving tomatoes or cucumbers and whether another curious item is an ice-cream server, a fish server, or a crumb scoop. The casual onlooker can easily wonder if anyone really knows what he or she is talking about.

  Suzanne MacLachlan is a collector of, among other things, any and all pieces of the Vintage pattern of silverplate, which was made from 1904 to 1918 by a division of the International Silver Company under the trademark “1847 Rogers Bros.” The pattern incorporated bunches of grapes into the handle design, and so collectors like MacLachlan, who at one time had eleven hundred pieces of Vintage, can with some reason call themselves Grape Nuts. An insurance agent’s request for an inventory of her collection forced MacLachlan to catalog her pieces, which led to the publication of her definitive Collectors’ Handbook for Grape Nuts. The book includes over sixty distinctly different pieces that she had actually seen and acquired, and it contains illustrations of another eighty or so items reproduced mostly from old silverware-dealer and jewelers’ catalogues. The pieces range from familiar dinner and salad forks to less common items such as marrow and cheese scoops. The distinctions among knife and fork and spoon can become so blurred that we find in the handbook one hybrid item identified as a “melon knife or fork” and another odd utensil designated an “olive fork or spoon.” This latter is a curious piece whose bowl has two vestigial tines and a large oval cutout in which one can imagine an olive easily resting, and it is a cataloguer’s challenge: “An early list has it only as an Olive Spoon, but later lists also include it in the fork column.” Other manufacturers were more directly ambiguous and called the patented piece an “Ideal Olive Fork and Spoon.”

  These three pieces of Vintage silverplate are, left to right, an “olive fork or spoon,” a chocolate muddler, and a tomato server. Had the olive retriever been made with a conventional bowl deep enough to hold the olive securely, a lot of undesirable liquid might have been transported to the plate. The annular bowl obviates this fault, and the vestigial tines serve to steady an olive before it is scooped up. Neither of the other two pieces would work for olives very well at all. (photo credit 8.1)

  In her preface, MacLachlan admits to the difficulties of trying to be definitive about the Vintage line: “every time (in two years of compiling notes) that I have written down a firm conclusion from experience, I have been promptly contradicted by additional information!” And later in the book she suggests a source of her difficulty: “A number of factors make grape collecting confusing—but fascinating. From 1904 to 1918 pieces were redesigned, renamed or possibly even changed in size. Special orders combined parts of different pieces, and manufacturers freely swapped blanks, inserts and brand names.” The problem is made quite explicit in the case of individual salad forks that carry identical catalogue numbers while having two distinctly different forms of tine. In the earlier model, the tines are wavy and come to rather sharp points; in the later model, the tines are straighter, shorter, stockier, and blunter. According to MacLachlan, this same fork also appears in various catalogues under the names “individual salad fork,” “individual pickle fork,” “pickle short fork,” and “individual relish fork.” She further notes:

  Tines on early models are often found badly bent. Tines on later models are straighter and heavier. Rogers’ 1847 line carried a lifetime guarantee, and the manufacturers upgraded any piece that required consistent repairs.

  The evolution of the fork’s design in response to the failure of earlier models to resist bending is a classic example of form following failure. That the evolutionary changes could take place within the fourteen-year lifetime of the Vintage pattern shows how responsive manufacturers can be to the failure of their products to function as intended. However, the change of the piece’s name, from salad to pickle to relish fork in the same short period of time, illustrates another, and much subtler, aspect of the evolution of form. Whereas bent tines can be judged by rather objective observation, this other aspect of formal evolution originates in a subjective perception of functional failure, not of the piece itself so much as of related pieces that it is intended to displace. Thus, the salad fork exists at all because the dinner fork somehow failed, or was thought or said to have failed, to function as an effective salad fork, perhaps by virtue of being too heavy or bulky for the lighter chore. The multiplicity of names for MacLachlan’s piece suggests its intended multiple uses, for which it was presumably better suited than other pieces.

  These forks represent some of the pieces of silverplate once available in the Moselle pattern, including, left to right, a pickle, pastry, and salad fork. The first two pieces show very pronounced cutting tines, which evolved from less robust tines that became bent when forks displaced knives in almost all aspects of eating. The unsymmetrical nature of these forks makes them distinctly right-handed. (photo credit 8.2)

  The problems faced by Grape Nuts are actually minor compared with what collectors of nineteenth-century patterns can face, since early-Victorian catalogues of silverware manufacturers were not even illustrated. Illustrations were common by the end of the century, however, perhaps because it was nearly impossible to tell some of the pieces apart—with or without a picture—or to identify corresponding pieces in different patterns. Between 1880 and 1900, Rogers Brothers introduced twenty-seven new flatware patterns, which included many ne
w kinds of serving pieces. Other silver manufacturers were equally prolific. According to Dorothy Rainwater, who has written widely on American and other silver:

  By 1898, the Towle Company’s “Georgian” pattern included 131 different pieces.… There were nineteen types of spoons for conveying food to the mouth, seventeen for serving, ten pieces for serving and carving, six ladles, and twenty-seven pieces for serving that were not classified as ladles, forks, or spoons. One can sympathize with the hostess of that day in trying to be sure that croquettes were not served with a patty server, or cucumbers with a tomato server.

  As late as 1926, some patterns were still being made with as many as 146 distinct kinds of utensils. To help simplify the situation for American industry, Herbert Hoover, then secretary of commerce, recommended—and members of the Sterling Silverware Manufacturers Association adopted—a list of fifty-five items as the greatest number of separate pieces that would be in any pattern thereafter introduced. Today it is rare to find more than twenty different pieces in a silver pattern. Confusion about the naming of specialized pieces remains, however, with current catalogues of the best silver companies still calling by disparate names items that look to serve identical functions. Thus, similar-looking pieces of silverware can be called a “cold meat fork” in one catalogue and a “cake or pastry fork” in another, or a “fish fork” and a “salad fork” in still other catalogues. The confusion may be compounded by the fact that differences in form between individual forks within a single pattern are often much less drastic than the differences between, say, what are called “dinner forks” in two distinct patterns.