His­tor­i­cally, fonts weren’t pro­tected by copy­right in the United States be­cause they were con­sid­ered use­ful items. That made sense, be­cause for most of that time, fonts were pri­mar­ily an in­dus­trial prod­uct, cast in metal by foundries.

In the ’60s, how­ever, metal type was eclipsed by type­set­ting sys­tems that of­fered eas­ily swap­pable fonts. These sys­tems al­lowed font de­sign to be sep­a­rated from font man­u­fac­tur­ing. This, in turn, al­lowed more com­pa­nies to put de­signs into the mar­ket­place. But un­der copy­right law, those de­signs still had no pro­tec­tion, spawn­ing a mar­ket for cheap (and le­gal) knockoffs.

In 1976, as the U.S. Con­gress pre­pared to amend the copy­right act, type com­pa­nies had an ob­vi­ous in­cen­tive to seek new pro­tec­tion for fonts. Luck­ily, these com­pa­nies found an in­flu­en­tial ally: the reg­is­ter of copy­rights, Bar­bara Ringer. Ringer tes­ti­fied to Con­gress that even if fonts didn’t de­serve full copy­right pro­tec­tion, they de­served more than noth­ing. To no avail—Con­gress didn’t en­act any new protections.

In the 1980s, the per­sonal com­puter ar­rived, and the shift to dig­i­tal fonts re­opened the ques­tion of copy­righta­bil­ity. Dig­i­tal-type ven­dors con­tin­ued to press their case with Con­gress, unsuccessfully.

But one of those ven­dors, Adobe, had a clever idea. To­day we know it as the com­pany be­hind Ac­ro­bat. But at the time, Adobe sold a large li­brary of dig­i­tal fonts. Rather than push for leg­isla­tive changes, Adobe en­cour­aged the Copy­right Of­fice to re­vise its reg­u­la­tions about the reg­is­tra­bil­ity of dig­i­tal fonts. The Copy­right Of­fice had al­ready rec­og­nized soft­ware pro­grams as copy­rightable items. If dig­i­tal fonts could be char­ac­ter­ized as soft­ware pro­grams, the think­ing went, then as­sumedly they should re­ceive pro­tec­tion too.

The fi­nal reg­u­la­tion is­sued by the Copy­right Of­fice in 1992 seemed to en­dorse this con­cept. The Copy­right Of­fice per­mit­ted fonts to be reg­is­tered as soft­ware, in recog­ni­tion of the fact that dig­i­tal fonts could po­ten­tially in­clude pro­tectable soft­ware in­struc­tions. Let­ter­forms them­selves, how­ever, would re­main un­copy­rightable. And no­tably, the scope of a copy­right in a par­tic­u­lar dig­i­tal font—which could po­ten­tially con­tain both copy­rightable and non-copy­rightable ma­te­r­ial—would re­main “a mat­ter for the courts to determine”.

That was the state of the law in 1997, when Adobe filed a copy­right-in­fringe­ment case against South­ern Soft­ware, a com­pany that had sold a lightly edited ver­sion of Adobe’s font li­brary. South­ern’s de­fense: the fonts were not copy­rightable, and thus no in­fringe­ment was pos­si­ble. The judge de­nied sum­mary judg­ment, rea­son­ing that the char­ac­ter shapes amounted to pro­tectable ex­pres­sion. The case set­tled shortly there­after, with no appeal.

Type de­sign­ers point to this de­ci­sion as con­fir­ma­tion that fonts are soft­ware and thus pro­tectable. Oth­ers have sug­gested that the de­ci­sion was an in­ac­cu­rate in­ter­pre­ta­tion of the 1992 regulation.

Was it? Maybe. But the ques­tion is largely ob­so­lete. Fonts are dig­i­tal files, and like every other dig­i­tal file, the in­ter­net will hap­pily pro­vide you a free copy. Yet the font busi­ness—com­pletely op­po­site to the mu­sic busi­ness in this re­gard—has con­tin­ued to grow and thrive in the in­ter­net era. There­fore, the font busi­ness must be based on some­thing closer to the honor sys­tem. It re­mains an un­usual coun­terex­am­ple to the idea that dig­i­tal rights man­age­ment and lit­i­ga­tion are in­dis­pens­able tools in in­tel­lec­tual-prop­erty economies.