Part 1 is here. An introduction and guide to my series of posts “Corpora and the Second Amendment” is available here. The corpus data that is discussed can be downloaded here. That link will take you to a shared folder in Dropbox. Important: Use the “Download” button at the top right of the screen.
Update: Concordance-line references have been changed to reflect revisions to the spreadsheet from which the lines were copied, as have figures for the total number of concordance lines and for the various subtotals that are given.
New URL for COFEA and COEME: https://lawcorpus.byu.edu.
In this post and the next one, I will discuss the corpus data for bear arms.
This post will focus on the data that I think is consistent (or at least arguably consistent) with the Supreme Court’s interpretation of bear arms in District of Columbia v. Heller, and the next one will deal with the data that I think is inconsistent with the Heller interpretation.
As I discussed in my last post, the court in Heller held that the “natural meaning” of bear arms in the late 18th century (i.e., its “ordinary meaning” (i.e., what it ordinarily meant)) was “wear, bear, or carry upon the person or in the clothing or in a pocket, for the purpose of being armed and ready for offensive or defensive action in a case of conflict with another person.” As I read the data, very little of it is consistent with that interpretation.
The data that I analyzed is from the BYU Law Corpora: COFEA (the Corpus of Founding Era American English) and COEME (the Corpus of Early Modern English). My searches were limited to texts from the period 1760–1799; COFEA is limited to that period, and although COEME covers a wider period, I restricted my search to the period covered by COFEA.
What I searched for were uses of arms (as a noun) within four words to the left or right any form of bear as a verb (i.e., bear, bears, bearing, bore, born, and borne). I downloaded all the concordance lines that were responsive to those searches. As is often the case, there was a lot of duplication within the results from each corpus, as well as overlap between the results of the two corpora. I removed duplicates from the results before doing any analysis, and after doing that 749 concordance lines remained.
Out of those 749 lines, I excluded an additional 218 lines from my analysis. These fell into two categories.
The first consisted of uses of bear arms as part of a phrase such as right to bear arms, right to keep and bear arms, and liberty (or privilege) of bearing arms—but only when the topic was the right to bear arms in the United States or in one of the U.S. states and the phrase occurred in the text (or proposed text) of a constitutional provision protecting that right. Given that the issue to be decided is how the right to keep and bear arms as used in the Second Amendment was likely to have be understood, there was nothing to be learned from considering uses of that very phrase or of closely related variants, in a similar context. Other uses of phrases such as the right to bear arms were not excluded from the analysis. Two of those uses were related to the United States, and after referring to the source documents in order to understand the relevant context, I categorized them as conveying the military sense. You can find these uses in concordance lines 502 and 503, with links to the source documents provided. The remaining uses of phrases such as the right to bear arms concerned the right as it existed at one time or another in other countries, and I categorized them as ambiguous. (Lines 525-527.)
The second category of uses excluded from the analysis consists of concordance lines that were linguistically irrelevant (e.g., bear arms was used in the sense relating to coats of arms, arms was used in its body-part sense, bear was used in its sense relating to childbirth, or bear and arms appeared within four words of one another but were not otherwise related.
Once all these concordance lines were set aside, 531 lines were left. That data can be found at Tab 2 of my spreadsheet for bear arms. (Note that in the examples from the corpus that are provided here and in my other posts, I have sometimes cleaned up the spelling and typographic or OCR anomalies.)
The data that is the subject of this post is in concordance lines 504–514 and 517-531. The remainder of the data will be discussed in my next post.
Lines 504-514 (representing only 11 concordance lines—barely 2% of the total) are the only lines in which bear arms, or a grammatical variant, such as “the arms they bore,” was unambiguously used to convey what would generally be thought of as its literal sense: ‘carry weapons.’ (My reason for hedging in calling that the literal sense of bear arms is explained at the end of this post.) This scarcity of literal uses is especially striking when it is compared against the fact that literal uses of carry arms accounted for more than half the uses of that phrase. (See Tab 3 of the spreadsheet for bear arms.)
This by itself is a grave problem for Heller’s interpretation of bear arms, because these 11 lines are the only ones in which bear arms could possibly have been unambiguously used to convey what Heller declared to be its natural meaning. After all, uses of bear arms that convey that meaning are by definition a subset of all literal uses. So any use that doesn’t clearly denote literal carrying can’t unambiguously convey the Heller-approved natural meaning.
And as if that were not bad enough, one of the 11 lines has to be excluded right away, because it unambiguously denotes the carrying of arms for a purpose other than being prepared for confrontation.
The source of this line is a biography of Captain John Smith. The context is that Smith has been captured by a group of Native Americans. Fearing that he will be killed, Smith tries to avoid that fate, or at least delay it, by making a gift of his compass to the Native Americans’ leader. Eventually, though, his captors “fastened him to a tree and prepared to dispatch him with their arrows.” But Smith’s life is suddenly spared:
(1) At this instant, the chief, holding up the compass, which he esteemed as a divinity, they laid aside their arms, and forming a military procession, led him in triumph to their village Orapaxe. The order of their march was thus: they ranged themselves in a single file, the King in the midst, before him were borne the arms taken from Smith and his companions; next after the King, came the prisoner, held by three stout savages; and on each side a file of six. [Concordance line 504; supplemented with extended context.]
It is clear from this text that the Native Americans’ purpose in carrying Smith’s weapons was not to use them, but rather to bring them to their village, and perhaps also to display them in the triumphal procession. So this use doesn’t convey the “natural meaning” that Heller endorses.
Once this concordance line is set aside, there remain only 10 lines concordance lines that both (a) clearly denote the physical act of carrying of weapons and (b) can be read as denoting that the weapons were carried for the purposes specified in Heller interpretation. They are as follows:
(2) Since custom has allowed persons of rank and gentlemen of the army to bear arms in time of peace, strict care should be taken that none but these should be allowed to wear swords [Line 504.]
(3) were observed, nor any aged persons, except one man whose head was bald, and he was the only one who bore no arms the other seemed to be select men, and rather under than above the middle age. [Line 506.]
(4) The company with which Dampier chose to associate, consisted of 44 men who bore arms, a Spanish Indian who bore arms also, two Moskito Indians, already described , with five slaves taken in [Line 507.]
(5) our little boy bears under his left breast the distinct mark of a bow and arrow, the arms born by one of these savages. [Line 508.]
(6) as strong as man, and equally ardent for women as for its own females; an ape who knows how to bear arms, to attack his enemies with stones, and to defend himself with clubs. [Line 509.]
(7) The uniforms of the troops, and the kind of arms they bore, confirmed to him the conjecture of Cavigni, and he had the ſatisfaction to ſee them paſs by, without even ſtopping [Line 510.]
(8) I fervently hoped that no new exigence would occur, compelling me to use the arms that I bore in my own defence. [Line 511.]
(9) Now, thick as hail, her fatal darts she sings; The two-edg’d ax now on their helmets rings. Her shoulders bore Diana’s arms and bow [Line 512.]
(10) a soldier from yon camp Fled to my post; hollow and gaunt he was; His shrivell’d limbs scarce bore his sounding arms. [Line 513.]
(11) When men high in command, men of fortune and family, fall, their deeds are blazoned, and they figure in history; but who, save the poor widow and the orphan, enquire after the very names of the rank and file? There they lie, a mass of human flesh, not so much regretted by the despots as the horses they rode, or the arms they bore. [Line 514; supplemented with extended context.]
Except perhaps for (6), which has to do with the bearing of arms by orangutans (no, I’m not joking, why do you ask?), it’s pretty easy to infer from the context that the arms in question were borne for purposes of confrontation, or at least in order to be prepared for confrontation. And because those inferences are based on the various contexts in which bear arms appears, they are analogous to the inference in (1) that the arms were not carried for purposes of confrontation.
This raises the question whether we’re actually looking at two distinct senses of bear arms. In other words, does the bear arms in (1) mean something different than the instances of bear arms in (2)–(11)?
If the interpretation in Heller is correct, the answer is yes: the natural or ordinary meaning is ‘carry weapons for purposes of confrontation,’ so ‘carry weapons for purposes of confrontation’ must constitute a separate meaning. But it’s not at all clear that that’s right. A different way of looking at the issue would be to say that bear arms is used in the same sense in all 11 of these examples, and that the inferences that are drawn about the purpose of bearing the arms are based on the phrase in combination with the co-text.
Think of it this way. Suppose that before Frank Muscarello left home to drive to the drug deal for which he was ultimately arrested and convicted, he had said to one of his criminal associates, “I’m leaving now to go sell some of this weed, and I’m taking my gun with me.” One could certainly infer from that statement that Muscarello’s purpose in taking the gun was to be prepared for confrontation, but I don’t think it follows that he was using the verb take any differently than if he’d said, “I’m taking my sunglasses.” It makes more sense to me to say that there would be nothing special about Muscarello’s use of take, and that any inference we draw about his purpose would arise from our knowledge of the situation as a whole.
If I’m right in what I’m arguing, these ten uses of bear arms provide no support at all for Heller’s interpretation. If I’m wrong, then they provide ten concordance-lines-worth of support, which is insignificant in comparison with the body of contrary data that will be the topic of my next post. And if there’s no way to decide for sure whether I’m right, which might well be the most likely possibility, these ten lines can’t be said to unambiguously convey meaning that Heller endorsed.
That would mean that the corpus data provides not a single example of bear arms being unambiguously used in that sense.
IN ADDITION TO the 11 concordance lines I’ve just discussed, there are 15 lines in which bear arms was not unambiguously used in its literal sense, but as to which I couldn’t point to a specific factor ruling out the possibility that they would have been understood to express the “natural meaning” that was declared by Heller.
(12) but know, Isidor, that even at this age, that man bears not arms in Spain from whose crest Rayo would not now, even〈 ◊ 〉old though he be, hope [Line 517.]
(13) Trade into our Channels. There are here a few White families supposed to be able to furnish about sixty arms bearing men who are said to be well disposed to our Government & who certainly in a controversy merely with the Indians [Line 518.]
(14) persons, under grievous penalties, to enter the palace: nor was it permitted tiem to go about th city, nor to bear arms. [Line 519.]
(15) good, and chastising the insolent disturbers of the peace of others. To this end was given them ample authority to bear arms, and to take with them armed men; to imprison delinquents, and accommodate all disputes which should arise [Line 520.]
(16) Why boasts he not his limbs grown black With bearing arms, or his strong back With which he threw the bar? [Line 521.]
(17) how unlike your negro ſlaves! When do you inſtruct them to read and write , and teach them to bear arms, to fight with their maſters? No, you are conſcious of guilt; therefore you muſt not inſtruct them, nor permit [Line 522.]
[Comments on (17): (a) I interpret “to fight with their maſters” to mean ‘fight alongside their masters,’ not ‘fight against their masters.’ (b) I initially categorized this line as ambiguous because of it appeared to me that if bear arms is interpreted as having its military sense, the inclusion of to fight would be at least somewhat redundant. However, it now occurs to me that if bear arms is interpreted to mean ‘serve in the military,’ there wouldn’t necessarily be any redundancy, since it is possible to do military service without actually having occasion to go into battle.]
(18) freedom of election to parliament, the freedom of speech in parliament, and the right of the subject to bear arms, and to petition his sovereign. [Line 523]
(19) Interpreter again, who then aſked of what Uſe can you be in any Kingdom or Government, ſeeing you will not bear Arms and fight? To this I replied, that many of us had borne Arms in Times paſt, and been in [Line 524.]
[Comment on (19): This line is analogous to (17), with the apparent redundancy being between bear Arms and fight. What I’ve said regarding (17) therefore applies here as well.]
(20) But when the members of communities were permitted to bear arms, and were trained to the use of these, this in some degree supplied the first defect, and [Line 525.]
(21) theſe towns into corporations; they did not not eſtabliſh a municipal government; they did not grant them the privilege of bearing arms. They contained nothing more than a manumiſion of the inhabitants from the yoke of ſervitude; an exemption from certain [Line 526.]
(22) the Roman Catholics, who are by far the most numerous, the free exercise of their religion, with the liberty of bearing arms, so long unjustly deprived of, and disarm in due time all the Protestants in their turn. [Line 527.]
(23) might only say, that dominion founded in arms, or rather in oaths, (for the most violent of the jurors never bore arms, except on muster-days) is as absurd as dominion founded in grace. [Line 528.]
[Comment on (23): I initially categorized this line as ambiguous because the reference to muster-days evoked for me the idea of literally carrying weapons, which raised the possibility that bear arms had been used in its literal sense. However, it now occurs to me that participating in musters could well have been regarded as performing military service, and that under that interpretation, bear arms would most likely have been used to convey that meaning.]
(24) the ſaid Hyrcanus, that the Jews, being tied up by the religion, laws, and conſtitutions of their country, not to bear arms, travel, or ſo much as provide themſelves neceſſaries for life, upon the day of their ſabbath, are thereby rendered [Line 529.]
(25) millions of people: So it was easy to levy twelve hundred thousand sword men in a country, where all people bore arms, and still to have corn to sell to strangers towards the purchase of cattle; for if the flocks that [Line 530.]
(26) were transplanted from a sect of christians in Pennsylvania. Perhaps those German sects of christians among us, who refuse to bear arms for the purpose of shedding human blood, may be preserved by divine providence, as the centre of a circle [Line 531.]
[Comment on (26): I initially categorized this line as ambiguous because I thought that if bear arms was interpreted as having its military sense, including for the purpose of shedding human blood appeared to me to be redundant, since that purpose seemed to be inherent in the idea of bearing arms and therefore would not have needed to be stated explicitly. However, if bearing arms is understood to include activities such as participating in musters, whose direct purpose was not to engage in violence (see the comment on (19), above), there wouldn’t necessarily be any redundancy.]
Other than my comments on specific lines, above, I don’t have much of interest to say about these uses. I think most of them probably convey the military sense of bear arms, but I’m not sure enough of that judgment to call them unambiguous. But even if these uses are read as instantiating the natural meaning described in Heller, they pale in significance when considered together with the data that I will discuss at in my next post.
“Literal” meaning, linguistic history, and carry arms
Earlier in this post, when talking about the sense of bear arms as meaning ‘carry weapons,’ I used a hedge when I referred to that sense as “what would generally be thought of as its literal sense.” I did that partly because the distinction between “literal” and “figurative” language is more complicated than you might think, but more importantly because the distinction is especially problematic in the case of bear arms.
We tend to think of literal meaning as having a kind of priority over figurative meaning; literal meanings seem more basic or central, with figurative meanings being derived from them. (From now on, I’ll leave out the scare quotes around literal and figurative.) But linguistic history and the corpus data make it pretty clear that whatever the validity of that view as to other words and expressions, it’s inaccurate in the case of bear arms.
As is mentioned in my last post, and discussed in more detail in my posts about bear and arms, it is a misleading oversimplification to think that the late-18th century meaning of those words was ‘carry’ and ‘weapons.’ By the time of the founding era, carry had replaced bear as the verb that was usually used to denote the carrying of tangible objects. In addition, although arms was still being used in its literal sense, it was also used in a wide variety of figurative senses that have by now largely fallen into disuse.
There is reason to think that the figurative uses of arms can be traced back to before the word became part of the English language. Arms came into English via Anglo-Norman (the version of Old French spoken in England after the Norman Invasion), and some of the figurative uses can be traced back to Anglo-Norman. One of those figurative uses was the porter armes (meaning ‘serve as a soldier, fight for a country or a cause’ (OED); literally, ‘carry weapons’), which was the predecessor of bear arms in its military sense. And the use of bear arms in that sense in English has been traced back to the beginning of the 14th century—almost as far back as the use of arms in any sense.
What all of this suggests is that the military sense of bear arms didn’t necessarily develop out of the literal use of bear arms in English. Rather, it was most likely absorbed into English as a full-blown idiom (with porter translated into bear) at roughly the same time as the literal use of arms. That conclusion is supported by the OED, which includes this etymological note in its entry for bear arms: “After Anglo-Norman and Old French, Middle French porter armes (c1100 in this sense; French porter armes).”
As the corpus data shows, bear arms was also used in the sense of ‘carry weapons,’ but the corpus data suggests that as the literal use of bear came to be replaced by the use of carry, the literal use of bear arms declined and was largely replaced by carry arms. As discussed earlier in this post (and as I’ll discuss further in the next post), the data for bear arms includes only a handful instances in which bear arms was used literally. In contrast, in the data for carry arms, more than half the uses are literal. And as is discussed in my post on bear, the pattern of usage of bear was by the second half of the 18th century quite different from the pattern for carry.
All of this is consistent with—and may well help to explain—the fact that the corpus data for bear arms is dominated by uses conveying the phrase’s idiomatic military meaning. (See Tab 2 of the spreadsheet for bear arms, and the discussion in the next post.) It seems reasonable to think that the persistence of bear in bear arms is attributable to the fact that the phrase was an idiom and therefore that its meaning was, by convention, associated with the phrase as a whole, rather than being derived compositionally from the separate meanings of bear and arms. Presumably the association of the phrase with the idiomatic meaning was stronger than the association of bear with the meaning ‘carry,’ which would have weakened over time as carry pushed bear out of the niche it had previously occupied. If, as seems likely, bear arms occupied a separate niche of its own, it would have been more or less unaffected by the competition between carry and bear—which was largely, but not completely, won by carry.
The strength of the association between bear arms and the idiomatic military meaning can be inferred from the corpus data showing that carry arms (Tab 3 of the spreadsheet) was sometimes used to express the same idiomatic meaning as bear arms. For example:
(27) He said he had carried arms some years for the king of France, his sovereign; after which he went into the Spanish army [Line 136.]
(28) Governors of places have often promised that, for a limited time, their garrisons should not carry arms against the enemy with whom they capitulated [Line 145.]
(29) Even in North and South-Carolina, where many of the tories carried arms, the whigs have shewed the magnanimity of conquest, as well as the spirit of liberty and wise government, by [Line 150.]
(30) I believe my countrymen, who last war carried arms, will candidly own, that had we no better reason for taking up arms than that just assigned, we [Line 154.]
(31) an absolute Tyranny is essential in the Government of an Army, and that every Man who carries Arms, from the General Officer to the private Centinel, must be content to be a temporary Slave [Line 155.]
If my reading of these concordance lines is correct, it seems likely that these idiomatic use of carry arms essentially amounted to uses of bear arms in which the old-fashioned bear was translated into the more up-to-date carry. That would have been possible because literal bear wasn’t entirely replaced by carry, as can be seen from the fact that the corpus data for bear weight and bear burden(s) included instances in which bear was used literally. So speakers of English would still have been aware of that use. That awareness, together with the strength of bear arms‘s association with the idiomatic military meaning, was presumably what led to the extension of that meaning to carry arms.
While it’s impossible to know for certain whether I’m right in what I think happened, it strikes me as the obvious explanation for the pattern of usage that the evidence reveals. And if I’m right, there would be converging lines of evidence supporting the conclusion that the military sense of bear arms was the one that was used most often and that was therefore its ordinary/natural meaning.
Update: The OED indicates that carry arms was used in a military sense as early as 1580; it also provides examples of that use from 1667 and 1784. But that’s not necessarily inconsistent with what I’ve said above. It’s entirely possible that the extension of the military sense from bear arms to carry arms is something that happened repeatedly over the years, if the use of carry arms in a military sense wasn’t frequent enough to become conventionalized. In that event, each use of carry arms in that sense would have represented an initial step toward a change in the language that never caught on. However, the data I’ve seen isn’t remotely sufficent for me to be comfortable trying to decide which scenario—conventionalization or independent reinvention—is more likely.
[Various changes have been made to this post since it first appeared, in addition to those noted in the update at the top of the post. The most significant are (1) an addition to the comment on example (17), and (2) the substantial reworking of the three paragraphs near the end of the post that begin with, “All of this is consistent with—and may well help to explain—the fact that the corpus data for bear arms is dominated by uses conveying the phrase’s idiomatic military meaning.”]
Cross-posted on Language Log.