Duke Center for Firearms Law
Duke Law logo

The ‘Strange’ Syntax of the Second Amendment

By on July 14, 2021 Categories: , ,

The Second Amendment is not sloppy or ungrammatical, as some modern analysts claim. Rather, the Amendment is written in a variety of English that no longer exists. Since none of us are native speakers of late 18th century American English, we cannot expect to have good intuitions about its grammaticality or interpretation. When we read Shakespeare, for example, we accept that we have to rely on footnotes about vocabulary and syntax. The language of the Bill of Rights is chronologically closer to Shakespeare’s English than to present-day English, so the words and syntax are not always going to be immediately comprehensible.

The Second Amendment seems especially confusing because its structure has been subject to syntactic change, not just changes to words or word meanings. Words change faster and more frequently than syntax, so they are easier to notice. As we get older, we notice young people using words in new and different ways, whereas we probably are not aware of many differences in their syntax. But when we look at Shakespeare’s English, for example, we can see that syntax does change. If something seems ungrammatical to us, that is a signal that we need to look at how the grammar was used by native speakers.

The Second Amendment consists of a subordinate clause, A well regulated Militia, being necessary to the security of a free State, followed by a main clause, the right of the people to keep and bear Arms, shall not be infringed. The main clause sounds perfectly grammatical in present-day English (if we ignore the extra comma, which does not seem to have been significant). The subordinate clause with being, however, seems to have something wrong with it. This is because the being-clause precedes the main clause, and the two clauses have different subjects. The last example of this type in the Corpus of Historical American English (COHA; a 475-million-word balanced corpus of American English 1820—2019) is from 1923. Since this type of being-clause fell into disuse around a hundred years ago, it is reasonable that modern readers would not have good intuitions about its grammaticality or its meaning.

In order to understand a syntactic construction that we no longer use, we have to look at historical examples of the construction. A good resource for this is a balanced historical corpus, that is, a collection of texts from a particular timespan and region that consists of a balanced mix of personal letters, newspapers, scientific treatises, religious texts and so forth. These corpora are intended to provide an idea of general usage at particular times and places. Relevant balanced corpora for studying the Second Amendment include COHA, mentioned previously, or A Representative Corpus of Historical English Registers (ARCHER; 3.3 million words, 1600—1999). These corpora can tell us how being-clauses changed over time. They can also show how being-clauses were used, so that modern readers have a better idea of how this grammatical construction shapes the meaning of the Second Amendment.

For the moment, let us consider only being-clauses with the basic structure of the Second Amendment, in which the being-clause precedes the main clause and has a different subject. (A wider range of being-clauses is examined in my paper “Being-clauses in historical corpora and the U.S. Second Amendment”.) Being-clauses of this type have had four possible meanings, several of which could overlap. First, they could signal that the event in the being-clause happened before the main clause event, as in this example from 1723: The morning being come and breakfast over, Stertorius’s coach was brought. That is, after morning came and breakfast was over, the coach was brought. This is called a temporal usage. It is the earliest kind of being-clause, and gave rise to the second and third types.

The second use of being-clauses, the conditional, has always been rare. Conditional being-clauses were used to make predictions, as in the following example from 1786: These things being granted what is of a like kind will readily be so disposed too. If the ‘things’ (atmospheric conditions) occur, then ‘what is of a like kind’ (condensation) is predicted to behave in a particular way. Temporal and conditional meanings can overlap, as in the last two clauses in this 1833 example: Increase the amount of Bank notes, and, other things being the same, prices will rise. Whenever other things are the same, this statement claims, prices will rise (a temporal meaning); and if things are the same, prices will rise (a conditional meaning). These clauses are related temporally and conditionally, and the temporal and conditional meanings are completely compatible.

The third use of being-clauses also evolved from temporals, and could likewise overlap with them. These being-clauses signalled real-world causation, as in this sentence from 1780: The usual passages for the waters below being obstructed, they flooded the low grounds. That is, flooding occurred because the passages were obstructed. This is an external causal because it refers to a cause and a consequence in the real world. It’s important to note that an external causal relation frequently assumes a temporal one. In the above example, the flooding happens because of the obstruction (a causal relation) but also happens at the time of the obstruction (a temporal relation), so both causal and temporal relations are present, and are compatible with each other.

The fourth type of meaning is an internal causal, where the being-clause provides the logical basis, not the real-world cause, for the main clause, as in this example from 1702: The words in the will being to Richard and the heirs of his body, the heirs were in that will only words of limitation, and not of purchase. Here, the being-clause gives the reason for concluding the status of the heirs stated in the main clause. The main clause could be paraphrased by it was concluded that the heirs were in that will only words of limitation, and not of purchase. An external causal can never be paraphrased this way, and it was concluded that the waters flooded the low grounds would make no sense in the external causal above.

If we assume that the Second Amendment was grammatical, then its being-clause belonged to one of these four types or a documented area of overlap between them. The temporal reading would indicate that whenever “A well regulated Militia” is “necessary to the security of a free State”, then “the right of the people to keep and bear Arms, shall not be infringed.” A conditional interpretation would entail that if “A well regulated Militia” is ever “necessary to the security of a free State”, then “the right of the people to keep and bear Arms, shall not be infringed.” The external causal interpretation would mean that “the right of the people to keep and bear Arms, shall not be infringed” for the purpose of “A well regulated Militia … necessary to the security of a free State”. The internal causal would indicate that because it is known that “A well regulated Militia” is “necessary to the security of a free State”, it is concluded that “the right of the people to keep and bear Arms, shall not be infringed”.

The temporal, external causal, and internal causal readings are not equally likely. The ARCHER corpus, for example, contains 37 being-clauses of the relevant type from the second half of the 18th century. Of these, 18 have purely temporal meanings without conditional or causal inferences; 1 is a conditional; 19 have external causal meanings; and there are no internal causals. Statistically, then, the temporal and external causal interpretations of the Second Amendment are the most probable.

In the context of the Second Amendment, these two interpretations are not incompatible. We have seen that external causal meanings often assume temporal ones, since effects usually happen along with their causes. Both a temporal and a causal reading would assert that “the right of the people to keep and bear Arms, shall not be infringed” whenever a militia was “necessary to the security of a free State”. The causal reading would additionally assert that the “right” was for the purpose of the necessary militia, and therefore applied whenever the militia was necessary.

A temporal or causal relation between the clauses would mean that the main-clause content was temporally or causally contingent on the being-clause content, and “the right of the people to keep and bear Arms, shall not be infringed” would only be asserted when, or for the purpose of, “A well regulated Militia being necessary to the security of a free State”. Interpreting the main clause while ignoring the being-clause would be nonsensical, and certainly contrary to the original intent or understanding of the two clauses.

Of course, understanding the relation between the clauses does nothing to answer the question of how often a “A well regulated Militia” was thought to be “necessary to the security of a free State” and consequently how often “the right of the people to keep and bear Arms, shall not be infringed”. Perhaps such a militia was thought to be a permanent necessity, in which case the right to bear arms for that purpose would be perpetual. This is an issue for historians, not linguists. It is also a matter of debate as to whether the founders’ opinions on this issue outweigh those of present-day theorists, who might disagree on the necessity of a “well regulated Militia” to the “security of the free State”. This is an issue for legal scholars. However, the history of being-clauses does suggest that the applicability of the Second Amendment’s main clause was temporally and potentially causally contingent on the necessity of the “well regulated Militia” to the “security of the free State”, which may delimit the reasonable range of debate for theorists in other areas. Recognition of this clausal relation may also be helpful for interpreting words and collocations that have been the subject of debate, such as militia and bear arms, since the context of these items would have played a role in disambiguating their meaning.

This linguistic history also makes it clear that the Second Amendment was grammatical and probably unambiguous at the time of its writing. The Bill of Rights, like Shakespeare’s plays, was not badly written. We can blame our incomprehension on the ongoing process of language change – and perhaps on present-day speakers who fail to acknowledge it.