Justice Scalia had strong opinions about dictionaries. While for many the best dictionary is whichever is closest on the shelf, not so for Scalia. No clearer was this on display than in his opinion in MCI v. AT&T, 512 U.S. 218 (1994). There, faced with a battle of the dictionaries—each side of the dispute wielding their own preferred interpretation of the word “modify”—Justice Scalia panned Webster’s Third, stating that it was “widely criticized for its portrayal of common error as proper usage.” This critique did not go unnoticed: Merriam-Webster’s editor in chief responded with barbs of his own in a letter to the New York Times language columnist William Safire.
That Justice Scalia felt a need to criticize a dictionary that offered a definition inconsistent with the meaning he determined to be correct may seem curmudgeonly, but to dismiss his critique in this manner would miss something important about his commitment to textualism. Because of the manner in which textualism privileges dictionaries as a key source of interpretive meaning, such disagreements between dictionaries undermine key tenets of textualism’s claim to legitimacy—predictability, neutrality and objectivity—as they force the interpreter to choose among seemingly authoritative sources.
But for the textualist, such choices are not limited to cases of dueling dictionaries. In a new article published in the NYU Law Review, “Textual Gerrymandering: The Eclipse of Republican Government in an Era of Statutory Populism,” Professors Bill Eskridge and Victoria Nourse argue that in every hard statutory interpretation case a Scalian textualist—a “new textualist,” to use their label—is presented with a series of choices: first, she must make a “choice of text,” that is, she must decide what statutory text is relevant to the matter at hand; second, she must make a “choice of context,” deciding what materials are relevant to confirm or clarify textual meaning. These choices present a problem for new textualists. If judges “gerrymander” their choices of text and context so as to achieve their preferred result, then the textualist judge is, to recall Judge Harold Leventhal’s quip about legislative history, doing the equivalent of “looking over a crowd and picking out your friends.”
Eskridge and Nourse offer an incisive analysis of this problem—saying far more than I could do justice to here. But of particular interest is their proposed solution: a new focus on what they refer to as “republican evidence,” meaning evidence of how the legislative authors actually used and understood the language at issue. Given the fact that the Constitution sets up a representative form of government, statutory text, they argue, should be understood in the context in which it was created—that is, the public record of legislative deliberations. To do otherwise, would be to allow “Congress and the idea of a republican government” to be “eclipsed by dictionary and canon-wielding judges.” To put it differently, by yielding to republican evidence, we let the Constitution pick the interpretive context, and not a judge’s seemingly esoteric views about whether the Oxford English Dictionary is more authoritative than Webster’s Third.
Holy Trinity v. United States Re-examined
To help explain how the use of republican evidence works in practice, Eskridge and Nourse were kind enough to refer to my own research reexamining a classic (even infamous) statutory interpretation case, Holy Trinity v. United States, 143 U.S. 457 (1892). The case concerned section 1 of the Alien Contract Labor Act of 1885, which barred employers from prepaying transportation to bring noncitizens to the United States to perform “labor or service of any kind.” The question presented was whether the statute covered a British rector who had contracted to work at the Church of the Holy Trinity in New York City. Although Justice Brewer, in his opinion for the Court, conceded that the minister may have fallen within the “letter of this section,” he nonetheless read the statute to exempt the pastor, relying in large part on the notion that in a “Christian nation” no one would have thought to apply such a law to a minister traveling from abroad.
In his Tanner Lectures, Justice Scalia sharply criticized Justice Brewer’s construal of the statute in order to vindicate some “unexpressed legislative intent.” According to Justice Scalia, it is clear that the pastor was engaged in “service” and thus fell within the scope of the statute—“end of case.” But how clear is it really? While it is true that the standard usage of the word “service” might well include a pastor, my impression on encountering the statutory language at issue in Holy Trinity was that it was likely a term of art carrying a specialized meaning in certain contexts—an impression supported by the appearance of the near identical doublet, “service or labor,” in the Constitution’s Fugitive Slave Clause. To test this hypothesis, I looked to the Congressional Record, seeing if I could discern any specialized meaning associated with the phrase from the manner in which the legislators actually used it.
To do so, using ProQuest Congressional, I searched the Congressional Record between the years 1884 and 1885 for the exact phrase “labor or service” (which also included “the labor or the service” because of ProQuest’s algorithm). My interest in conducting this search was to discern how the legislators understood the language at issue—not to see how the legislators subjectively “intended” the statute to be applied. To avoid mixing up these two, I looked only at those instances of the phrase that arose in contexts not connected to the debate on the Alien Contract Labor Act. That limited my examination to three legislative debates in which the phrase appeared.
The result was surprising. Upon reviewing the Congressional Record, it became evident that the phrase “labor or service” was associated exclusively with work compensated by wages, rather than salaries. Moreover, based on other evidence I examined, in the late nineteenth-century there was an understanding that certain types of employees, such as laborers, would receive wages, while others, such as clergymen, would receive salaries. Indeed, newspaper accounts at the time make clear that Dr. Warren, the pastor at the center of the dispute in Holy Trinity, was to receive a salary. And there it was. Holy Trinity was, on textualist grounds, correctly decided—contrary to Justice Scalia’s famous criticism.
Although Eskridge and Nourse discuss my conclusion, they do not include the analysis itself. So, to help illustrate how “republican evidence” can help resolve tough cases—indeed, even a case that stirs up interpretive debate a century later—I detail below the three instances in the Congressional Record that came up in my search and which suggest, in my view, that “labor or service” was a term of art used to refer to wage labor.
- Statement by Representative Dingley: April 30, 1884
The first result appeared in an extended address by Representative Dingley of Maine on April 30, 1884. (15 Cong. Rec. 3588 (1884).) Discussing a bill to lower tariffs, Dingley argued that although reducing the protective duties would lower prices, it would lower wages even more and—as a result—cause the relative price of goods to go up. In making this argument, he stated that “most men really ultimately pay for what they buy in labor or services, and the practical question is not whether the money cost is more under protection than under free trade, but whether the cost in labor or service or some other product is more.” To illustrate this point, he offers a chart showing the change in wages between 1860 and 1880 for six types of manufacturers (wool, cotton, iron, machine, paper-making, and boot and shoe).
One inference to draw from this is that “labor or service” was, as a phrase, associated with industrial workers and wage earners. The link between the phrase “labor or service” and wage-earning in manufacturing is strengthened when Dingley next addressed the effect of the tariff on farmers, beginning with the statement: “every gentleman who has spoken in opposition to protection has assumed that notwithstanding protective duties may benefit the laborer and many other classes, yet that they impose a serious tax on every farmer without corresponding benefit.” He then stated that lowering duties would harm farmers because laborers who lose their jobs would become farmers leading to an increase in agricultural production and correspondent reduction in prices.
Another observation is that although Dingley used the phrase “labor or service” initially, in the next paragraph he instead referred simply to “labor.” In contrast, he never used the word “service” alone. This suggests that the phrase “labor or service” is specific legal term with a meaning distinct, at least as compared to “service” alone.
- Statement of Senator George: January 9, 1885
The second occurrence of the phrase was in a speech given by Senator George of Mississippi on January 9, 1885. (16 Cong. Rec. 575 (1885)). The context was a bill to regulate the freight fees that could be charged by railroads. Senator George pressed that the bill should prohibit the charging of discriminatory fees based on the amount shipped; that is, he wanted the bill to prohibit the practice by which railroads would charge a lower per-item cost for larger shipments. He said this was necessary in order to prevent the greater “centralization or concentration in commercial and industrial affairs.”
In describing the “evils” he noted how concentration undermines “independent workers in independent employment,” causing them to be absorbed into “labor workshops under a master.” But, he also stated that it would be “despotic” to try to counter this by directly regulating the prices of goods or “the price which one man may charge another for work or labor.” He then went on to say: “If we cannot fix the price of commodities as between private traders nor regulate the price of labor or service in private employment, we can and do accomplish this to a great extent in public employments.”
Read in context, it seems apparent that when utilizing the phrase “labor or service,” Senator George had in mind the wage-earning laborer. More particularly, it seems he was primarily referring to those working in the industrial and commercial sectors. This aligns with the association implicit in Representative Dingley’s statements discussed above, namely the association between “labor or service” and wage-earning in the manufacturing sector.
- Statement of Senator Bayard: April 17, 1884
Broadening the search to include “the labor or the service” adds one additional result from this period: a comment by Senator Bayard of Delaware on April 17, 1884 in the context of debate over a bill to establish a federal bankruptcy system. (15 Cong. Rec. 3049 (1884)). Senator George proposed an amendment to a portion of the bill giving a priority claim, above the other creditors, for “Wages due any workman, clerk, or servant of the bankrupt for labor, not exceeding $100.” Arguing that the $100 was too low, he proposed striking it altogether, ensuring a “laborer . . . be paid his full debt” if his employer went bankrupt.
Senator Bayard approved of the amendment, suggesting it was only fair given that “the very value of the assets . . . might be chiefly or in a large measure owing to the labor or the service of the persons whose interests we are not considering.” But he raised the concern that courts would interpret the word “servant” broadly to include even those with high salaries—specifically distinguishing a “salary” from a “wage.”
Senator Ingalls then interjected to agree that the term “servant” might be understood to include “the confidential manager of a great manufacturing corporation, or of a railroad, or of any other greater commercial enterprise, having a salary far in excess of that paid to members of this body.” He then explained that he understood the design of the amendment “to permit the payment of wages earned by manual labor,” and stated that he supported the amendment were it modified to effect this goal. Interestingly, Senator Bayard responded to this by stating “The Senator surely would not exclude other than manual labor. Clerical labor is just as hard and just as valuable.”
Looking at Senator Bayard’s comments as a whole, it appears that he understood “the service or the labor” to refer broadly to wage labor. Further, while it appears from context that the terms were associated with industrial work, the exchange with Senator Ingalls is a strong indication that the phrase referred to a category of work that was broader than manual labor. In particular, it appears that the primary characteristic distinguishing the category of “labor or service” from other types of work was whether compensation was paid in the form of salary or wages. This is consistent with both of the uses discussed above, as the Members of Congress in those instances used the phrase to refer to wage laborers.
Wage vs. Salary in Context
Given the way in which the legislators used the phrase “labor or service” in other debates during the relevant period, it is likely that as used in the Alien Contract Labor Act of 1885, the phrase is a term of art used to refer to wage labor—and not the sort of work typically compensated through a salary. The definition of “salary” in the 1828 edition of Webster’s Dictionary further supports this interpretation and, indeed, helps resolve the dispute at the heart of Holy Trinity. Webster’s defined “salary” as follows:
The recompense or consideration stipulated to be paid to a person for services, usually a fixed sum to be paid by the year, as to governors, magistrates, settled clergymen, instructors of seminaries, or other officers, civil or ecclesiastical. When wages are stated or stipulated by the month, week or day, we do not call the compensation salary but pay or wages; as in the case of military men and laborers.
As the definition quoted above makes plain, while “wages” were associated with “military men and laborers,” a “salary” was associated with “settled clergymen” among others. (Apparently, a “settled clergymen” was one who was hired by a church on a long-term basis.)
This suggests that Dr. Warren, the pastor at the heart of the Holy Trinity case, was salaried. Indeed, according to newspaper reporting at the time, Dr. Warren was to receive a set salary from the church. Taking these pieces of evidence as a whole, we could then conclude that “labor or service” referred to the sort of work which was typically paid through a non-salaried wage, excluding clergymen—Dr. Warren included.
Of course, this need not be the end of the analysis. In order to make exceedingly clear that my search was not for subjective “legislative intent,” I excluded from my analysis the actual debates over the 1885 Act itself. Eskridge and Nourse expanded the analysis to the 1885 debates and found that congressmen on both sides used “labor” and the phrase “labor or service” to refer to manual labor—which is broadly consistent with my conclusion above. I also refer the reader to Eskridge and Nourse’s analysis of the structure of the statute, which—they suggest—comes into focus once read in the context of the republican evidence cited above. Here then, we can see the statutory interpretation process at its best—different sources of interpretive meaning reinforcing each other and exemplifying the hermeneutical approach to statutory interpretation that Bill Eskridge has, for decades, advocated.
More broadly, thinking of the Congressional Record as a “dictionary” of sorts helps explain the value of legislative materials in a manner that sidesteps conceptual objections to the search for legislative intent. To an extent, this is not an innovation. Over a decade ago, Justice Thomas explained as much in his concurrence in McDonald v. Chicago, 561 U.S. 742 (2010). There, he noted that “[s]tatements by legislators can assist in [the interpretive] process to the extent they demonstrate the manner in which the public used or understood a particular word or phrase.”
What Eskridge and Nourse add to this idea is an important point about judicial humility in our constitutional structure: by centering the meaning of the language at issue as demonstrated by legislators’ deliberations, courts can avoid displacing the legitimate policy decisions of the most representative branch while still engaging in an interpretive process that can appropriately be labeled textualist. At a moment when many are concerned about challenges to the legitimacy of the Supreme Court—and by extension, the entire federal judiciary—this is a valuable perspective.
Daniel B. Listwa (YLS ‘19) is a law clerk to the Honorable Jed S. Rakoff, U.S. District Court Judge of the Southern District of New York.
 Id. at 3589.
 Id. at 3051.
 See Richard D. Brown, Knowledge is Power: The Diffusion of Information in Early America, 1700-1865, at 66 (1989).