Safe Harbor Originalism

safe harbor

Originalism’s success has resulted in a rapidly expanding body of scholarship by a richly diverse group of constitutional theorists, many of whom “tweak” the method in order to bring it within their preferred normative theory. This is the cost of success—everyone wants to play.

There are many ways to be an “originalist.” However, not all ways are originalist, and even those that are arguably originalist will not be equally accepted by practitioners of the method. If originalism is to maintain a degree of coherence as an interpretive option, its advocates are now pressed to define it, and to do so in a manner that distinguishes the method from its rivals while still leaving room for healthy exploration, disagreement, and development.

Thanks to the work of Professor Lawrence Solum, we have an easily understood, robustly theorized, and broadly accepted definition. Originalism claims that 1) the meaning of a text is fixed at the time of its adoption and 2) this constrains judicial application of the text. There may be other ways to define originalism, but any work that accepts Solum’s definition will be accepted as “originalist” by most academics, lawyers, and judges.

How does one go about determining the textual meaning that was fixed at the time of adoption? Once again, different scholars have adopted different methods. What follows is an originalist methodology that I believe all originalists accept as a properly originalist methodology. It has four basic steps, and to fall within the safe harbor, all four must be completed, and completed in order.

Step 1: Identify a Text

It seems fairly obvious that originalist arguments about the meaning of constitutional text should begin by identifying the text to be explored. Surprisingly, however, scholars frequently place this essential step at the end of their argument or skip it altogether.

For example, a great many putatively originalist arguments use the following form:

The Fourteenth Amendment originally was understood as protecting x, therefore the only question is whether x can be understood as requiring y.

The problem with this argument is that it does not tell us which text of the Fourteenth Amendment was understood as protecting x. The political exclusion of former rebels (Section Three)? The formulae for determining congressional representation (Section Two)?

Fourteenth Amendment scholars, of course, generally focus on Section One of the Amendment, but this opening section also contains several different provisions with clearly different meanings (citizenship, privileges or immunities, due process, equal protection). Safe harbor originalism requires the writer to identify the specific text that establishes a particular meaning.

Occasionally, scholars are unable to find a particular meaning linked to one specific text, and so attempt to link a claimed “original meaning” to a combination of texts or the Amendment in general. However, until they engage the original meaning of specific texts, they have not made a broadly acceptable originalist argument about any particular text. More likely, they have identified evidence of general historical expectations, as opposed to claims about the original meaning of constitutional text.

For example, consider the following claim:

Whether located in the equal protection clause or the due process clause (or privileges or immunities clause), it is clear that the Framers of the Fourteenth Amendment believed its adoption would accomplish x.

The scholar then provides historical examples of Framers’ expectations. Evidence regarding original expectations may be quite relevant to determining the original meaning of the text. But unless that evidence is part of an effort to identify the meaning of a particular text, “global expectations” arguments are less likely to be accepted as arguments about the original meaning of a constitutional text.

Indeed, some originalists reject altogether evidence of expectations. Even if other originalists are open to evidence of original expectations, if that evidence is not tied to a particular text, it is less likely to be broadly accepted as evidence of original textual meaning.

The same is true for arguing that a “synthesis” or combination of constitutional texts establishes a textual meaning that was not established by either of the individual texts on their own. For example:

The Fourteenth Amendment, when synthesized with the Nineteenth Amendment, establishes the constitutional right to x.

Whatever the merits of this approach, “synthesis” arguments cannot escape the need for establishing the original meaning of each of the synthesized texts—not, at least, if they wish to be broadly accepted as originalist arguments. If the meaning is not an identifiable part of the original meaning of either text, neither is the synthesized meaning.

Bringing in expectations or synthesizing meaning are two ways of avoiding engaging the text. Another is to rely on general assumptions about the text. For example:

It is commonly accepted that the original meaning of the Equal Protection Clause contained an anti-caste principle. Therefore, if we can locate x within that anti-caste principle, we have established x as part of the original meaning of the Fourteenth Amendment.

No text is cited. Instead, the argument relies on what the author believes is a consensus understanding of the text. I want to be careful here: All scholars can and should rely on the work of others. This is how we make progress. Still, arguments that rely on common consensus are not likely to be viewed as distinctly originalist. It may be, for example, that common consensus significantly diverges from the actual original meaning of the text, and the fact that we think otherwise has to do with errors in prior scholarship (or our erroneous understanding of the prior work).

This possibility is especially important to bear in mind given the development of originalist theory over the past two decades. Scholars who rely on “originalist” scholars from decades in the past run the risk of undermining the acceptance of their work if that earlier scholarship was based on a poor or incompletely executed form of originalism. Indeed, there are articles in my past that I would do differently (and hopefully better), if I did them now.

In sum, the first step of a broadly accepted originalist argument is to focus on a particular text and explore the original meaning of that particular text.

Step 2: Identify Patterns of Usage

Originalism seeks the original meaning of a text (see Step One). Determining that meaning requires an investigation of how the text was likely to be understood at the time of its original creation and adoption. The same text, after all, can have different meanings depending on the historical context in which it was first communicated. As we are often reminded, the term “domestic violence” most often communicates a meaning today quite different from what it conveyed in 1787.

In Step One, the originalist scholar identifies the textual terms to be explored. In Step Two, the originalist seeks examples of the actual usage of those terms at the time of framing and adoption. Step Two thus accepts what Solum calls the Fixation Thesis, which claims that original meaning is fixed at the time of the original adoption of the text.

There are, of course, many forms of historical evidence besides patterns of usage at the time of adoption that may be relevant to determining the original meaning of the text. Sometimes, it can be helpful to understand the deep background of legal terms, perhaps by exploring English common law, Blackstone’s Commentaries or the work of John Locke. Other times it may be illuminating to consider post-enactment discussion and case law. However, since the effort is to discern the meaning at or near the time of enactment, the most broadly accepted form of evidence involves patterns of usage at or near the time of enactment.

Because a single example might reflect idiosyncratic understandings, the originalist seeks to identify patterns of usage of sufficient number and clarity as to allow a reasonable conclusion about the commonly accepted meaning of the text at the time of adoption. For example, in the period 1866 to 1868, historians have identified numerous examples of the term “privileges and immunities of American citizenship” being used in a manner that expressly included freedom of speech and assembly. These references in the historical record are so common that they have convinced a broad range of originalist scholars that, at the very least, the Fourteenth Amendment’s words, “privileges or immunities of citizens of the United States,” referred to the citizen’s right to freedom of expression.

Occasionally, scholarship purporting to be originalist avoids discussing common usage of a particular text and instead focuses on historical evidence of common ideas or beliefs or expectations that were in play at the time of a text’s adoption. For example:

Most members of the 39th Congress believed in natural rights.

Unless the scholar provides evidence that such ideas, beliefs, or expectations were part of the original meaning of a particular text, the argument fails the second step of safe harbor originalism. As a result, fewer scholars, lawyers, and courts are likely to accept such generalized evidence of “ideas” or “expectations” as actually establishing the likely original meaning of a particular text.

A word must be said here about context. Identifying the original meaning of a text involves more than just identifying one or more dictionary meanings, or establishing how the rules of English grammar control a string of words. Originalists commonly accept the idea that the original meaning of the text includes the communicative meaning of the text. Words communicate different meanings depending on the historical and cultural context in which they are delivered.

So, for example, the words “equal protection of the law” if used today may communicate a very different meaning than that communicated by the same words in 1866. Once again, identifying patterns of usage helps establish the communicative meaning of a text communicated in a particular historical context. As we shall see in Step 3, establishing context also assists in determining whether a text communicated a broadly expandable concept or a more linguistically confined conception. (Fine work by Solum—responding to Dworkin—on this distinction.)

Finally, by requiring the scholar to identify a pattern of usage, as opposed to a single example of usage, Step 2 seeks to avoid the problem of “cherry-picking” historical evidence. It is not enough show that James Madison or John Bingham held a particular view of the text. To put forward what will broadly be accepted as an originalist argument, one must establish a common pattern among members of the relevant linguistic community.

Step 3: Identify the Floor and Ceiling of Original Meaning

Originalists arguments rarely establish, or claim to establish, the full and definitive meaning of a particular text. It is rarely the case that they need to. Most often, it is enough to simply resolve an ambiguity (does “domestic violence” refer to spousal abuse or civil insurrections?). Other times, it is enough to reduce the degree of vagueness surrounding a given text. For example, in seeking to determine whether a claimed right is part of the Fourteenth Amendment’s Privileges or Immunities Clause, it might be enough to determine that a particular right clearly was (or clearly was not) a part of that original meaning, without having to establish comprehensive definition of “Privileges or Immunities.”

Resolving ambiguity and reducing vagueness can be seen as establishing boundaries of possible original meaning—what I call establishing the “floor” and the “ceiling” of textual meaning.

Establishing the floor of original meaning involves identifying what must have been included as part of the text’s original meaning. Establishing the ceiling involves identifying meanings that cannot have been part of the original meaning. Between the floor and ceiling there is room for debate, or what some scholars call “construction.” In this in-between space, any construction of textual meaning that includes the floor but does not breach the ceiling is permissible, though not required by the available historical evidence.

Sometimes it is enough for the argument at hand to establish a floor. Other times, a ceiling is necessary as well. For example, scholars arguing that the Constitution supports a right to do x may try to build upon an established floor and amplify the scope of textual meaning by describing the identified right at ever higher levels of generality. These arguments may purport to be “originalist” in the sense that the constructed meaning is claimed to be consistent with the established floor or minimum original meaning. So, for example, a scholar might argue that the right to same sex marriage is consistent with the established minimal original meaning of the Equal Protection Clause, using as support a broad understanding of the term “equal protection.”

However, if the argument is to count as a broadly acceptable originalist argument, we must know whether construction is consistent with both the floor and the ceiling of a given text. Words like “equal protection” and “privileges or immunities” can be used to convey a broad concept with little, if any, fixed ceiling of meaning. On the other hand, these terms can also be used to communicate a particular conception of equality or liberty, meanings with historically identifiable floors and ceilings.

For example, I have argued that the members of the 39th Congress carefully excluded language that might be understood as nationalizing common law civil rights. Whatever else “privileges or immunities of citizens of the United States” included, it was understood as leaving common law civil rights to the states so long as states provided due process and equal protection. I argue, in other words, that the Privileges or Immunities Clause had a ceiling.

Of course, I might be wrong. The issue remains a matter of historical investigation and there remains much work to be done. The point is that claims about a permissible construction of a text will not be broadly accepted as originalist unless the construction fits within an established floor and ceiling of original meaning.

Step 4: Identifying the Construction Zone

As is clear from Step 3, establishing the floor and ceiling of original meaning often has the added benefit of identifying a remaining range of permitted, though not required, construction. By construction, I mean the application of the text to resolve a legal controversy in a manner consistent with, but not required by, the original meaning of the text.

Most originalists accept the existence of a “construction zone”—the place where choices of application are consistent with, but not demanded by, the original meaning of the text. The work of Jack Balkin, for example, carefully distinguishes the author’s claims about the original meaning of the text from textually permitted constructions. Although Professor Balkin has become something of a bête noir among certain kinds of originalists, many of those critics fail to recognize how much of Balkin’s claims about contemporary application involve choices among possible constructions, and not historical claims about original meaning. Other originalists such as Lawrence Solum and Randy Barnett make the same distinction between original meaning and permitted, but not required, construction. To these scholars, how one chooses among permitted constructions depends on one’s underlying normative theory.

Although most scholars and judges accept the distinction between original meaning and construction, a problem arises when they skip to Step 4 without having completed the first three. For example:

Whatever the original understanding of the Due Process Clause, nothing prevents later generations from constructing their own understanding of the majestic generality of constitutional liberty.

The problem with this argument is that it explores possible constructions without having first explored the text, its original patterns of usage, and its likely “floor” and “ceiling” of meaning. Absent these critical steps, we cannot know if the original meaning of the term was a majestic generality (a concept) or a specific term of art (a conception). Constructions that proceed without the necessary historical work are not likely to be accepted as consistent with originalist theory.

I should note that some originalist scholars deny the existence of a non-originalist construction zone. Michael Rappaport and John McGinnis, for example, argue that the method of choosing among permitted constructions is itself part of the original meaning of the Constitution. Similarly, Michael Paulson and Gary Lawson claim that applying a set of default rules in cases of underdetermined textual meaning can eliminate the need for construction. What McGinnis and Rappaport call “original methods originalism” may or may not be an aspect of the original meaning of the text (I have argued that it is not).

Be that as it may, what is important is that the scholar acknowledge and identify where original meaning ends, and construction begins. As our historical knowledge increases, the range of uncertain yet permissible constructions will likely decrease. Most originalists understand, however, that historical inquiry cannot fully resolve every issue of judicial application of the text.


Lest I be accused of excluding everyone from the safe harbor except myself, let me confess that much of my own early work does not qualify, while much contemporary work with which I disagree does qualify. In fact, as far as I can tell, contemporary originalists of almost all stripes currently accept the “safety” of the above outlined steps and generally endeavor to work within them. This includes scholars with whom I have serious disagreements regarding the meaning and scope of provisions such as the Ninth Amendment, the Eleventh Amendment, and the Privileges or Immunities Clause of the Fourteenth Amendment.

In fact, the current agreement regarding accepted methodology, despite lively disagreements regarding meaning, prompted this guide for the perplexed. In identifying the area of overlapping consensus, I hope to have helped those who regularly produce originalist scholarship to identify and more clearly explain our respective departures from common method.

To the extent that we depart, we should accept the burden of justifying our departure. We should also accept the possibility that our conclusions may be less persuasive to those who choose to remain safely in the lee of the island.