Carrying something for protection

Some people, after being frightened, followed or attacked, think about carrying something next time: a spray, a knife, a heavy torch, and in the US, for some, a gun. This lesson sets out what the law says about that in England and Wales, in Scotland and in three US states, and what researchers have found, and still argue about, when they study guns kept for protection. It gives no advice on choosing, buying, carrying or using anything, and it names no product. It gives you the rules and the evidence, so that a decision you make, or a claim you hear, starts from what they actually say.

Three limits first. Every legal rule here comes with the course's usual caution: this is general information, not legal advice. The law differs by state, and between England and Wales, Scotland and Northern Ireland. Nothing in the lesson tells you how to use anything, or what to choose, store or carry, guns included. And the second half, on the research about guns, is a contested question: the course sets out each side in its own words and takes none.

Why "for self-defence" is the problem

This is general information, not legal advice. The law differs by state, and between England and Wales, Scotland and Northern Ireland.

In England and Wales the main law is one sentence of the Prevention of Crime Act 1953, section 1: "Any person who without lawful authority or reasonable excuse, the proof whereof shall lie on him, has with him in any public place any offensive weapon shall be guilty of an offence".¹ The weight of the section is in what counts as an offensive weapon: "any article made or adapted for use for causing injury to the person, or intended by the person having it with him for such use by him or by some other person."¹

Read that definition as three routes. An article can be made to injure, adapted to injure, or neither of those and carried with the intention of injuring someone.

The Crown Prosecution Service (CPS), which prosecutes in England and Wales, reads the section the same way in its guidance for prosecutors, and adds, in brackets, that the intention in the third kind can be "either offensively or defensively".³ That is the mechanism of the whole section, and it is why "for self-defence" is the problem rather than the answer. Carrying something so you can hurt an attacker is the intention the third route describes. The statute defines the offence by what the article is for; it does not say why Parliament chose that, and this course read nothing that does.

Two further parts of the sentence matter. The offence has a defence of lawful authority or reasonable excuse, and "the proof whereof shall lie on him": it is for the person charged to show the excuse, not for the prosecution to disprove it.¹ And the CPS says how fear fits in:

Carrying an offensive weapon in public merely because of the possibility of a constant or enduring threat (even where genuine) does not amount to self-defence or a reasonable excuse. Exceptionally, it may be reasonable if the offender is in anticipation of an imminent and specific attack, such that the possession is closely connected in time and purpose to that immediate threat, and the weapon is being carried for the offenders own personal defence[3]

The guidance cites a 1972 case for that rule. Notice the words "(even where genuine)". The rule does not doubt that the fear is real. It says a real, continuing fear is not enough, and keeps the exception for an attack that is both imminent and specific.

One distinction matters here. The carrying offence is about having something with you. The CPS notes that it does not reach someone who picks up and uses an object on the spot; it covers people who have the weapon with them.³ Using something that comes to hand when you're attacked is judged under the law of self-defence in lesson 13, "The law on using force", by the questions of necessity, proportion, imminence and belief. Deciding in advance to carry something to hit with is a different act, and it meets this section first.

Blades

Knives have their own section, 139 of the Criminal Justice Act 1988, and it works differently. It covers "any article which has a blade or is sharply pointed except a folding pocketknife", and a folding pocketknife comes back in "if the cutting edge of its blade exceeds 3 inches".² Under this section there is no need for any intention to injure: the CPS says a blade does not have to be carried with any offensive intent for the offence to be made out.³

Two details catch people. "A 'lock-knife' does not come into the category of 'folding pocketknife'", in the CPS's words, so a locking blade under 3 inches is still covered.³ And the defences are specific. The section allows "good reason or lawful authority", and names three: having the article "(a) for use at work; (b) for religious reasons; or (c) as part of any national costume".² The CPS adds that having a blade for work is not automatically a good reason: prosecutors look for a close link between the article and the work, and may ask why a less dangerous tool was not used.³ Both this offence and the offensive-weapon offence carry up to four years' imprisonment when tried in the Crown Court (on indictment).¹²

A folding pocketknife with a blade of 3 inches or less is outside section 139, as long as it does not lock. It is not outside the 1953 Act: if it is carried to use on a person, the third route applies to it as to anything else. And some folding knives are an offence to have with you whatever their length: the CPS lists having a flick knife or a gravity knife with you as an offence under a separate Act, the Restriction of Offensive Weapons Act 1959.³

Scotland: the same shape

Scotland has its own statute, the Criminal Law (Consolidation) (Scotland) Act 1995, with section 47 on offensive weapons and section 49 on blades and points.⁴ Its definition of an offensive weapon has the same routes, including an article intended by the person who has it for causing injury, and both sections give a defence of "reasonable excuse or lawful authority".⁴ Section 49 has the same three named defences (work, religion, national costume) and the same 3-inch folding pocketknife, given there as 7.62 centimetres.⁴ What this course did not read is any Scottish prosecutors' guidance, or any Scottish case, on whether fear of attack can be a reasonable excuse. So it cannot tell you how a Scottish court treats that question, only that the statute puts it.

Pepper spray: a prohibited weapon in Great Britain

Sprays are not an offensive-weapons question at all. Section 5 of the Firearms Act 1968 lists weapons subject to a general prohibition, and paragraph (1)(b) includes "any weapon of whatever description designed or adapted for the discharge of any noxious liquid, gas or other thing".⁵ The government's border guidance counts "CS gas canisters, pepper sprays and other self defence sprays" among firearms.²² So pepper spray and CS spray, a kind of tear gas, are prohibited weapons, and the section extends to England, Wales and Scotland.⁵ Three consequences follow, and they are where most misunderstandings sit:

• Possession is the offence, not only carrying: a spray kept in a drawer at home is covered.
• Purpose makes no difference. There is no self-defence version of a prohibited weapon.
• It's a firearms offence, under the same section as automatic weapons.

This course did not read the penalty provisions, so it gives no penalty. Lesson 6, "Out and away", gave the short answer for Great Britain when it quoted the US State Department's advice to check the law before taking a spray abroad; this is the law behind it. Some shops sell sprays described as legal alternatives. This course read nothing on their legal status and names none.

This is general information, not legal advice. The law differs by state, and between England and Wales, Scotland and Northern Ireland.

Northern Ireland: different again

Section 139 has a separate Northern Ireland version, and Northern Ireland has its own firearms law, which this course did not read.² So it says nothing about sprays or offensive weapons there, beyond this: do not assume the rules above apply.

In the US, state by state

This is general information, not legal advice. The law differs by state, and between England and Wales, Scotland and Northern Ireland.

The US has no single rule. This course read three states' statutes on sprays, chosen because they differ, and nothing on other weapons.

• California, Penal Code §22810, is written about "tear gas", which the code defines to include "any liquid, gaseous or solid substance intended to produce temporary physical discomfort or permanent injury through being vaporized or otherwise dispersed in the air".⁶ It bars people convicted of a felony or of any crime involving an assault, people addicted to any narcotic drug, and minors from having one (with an exception in the next section, §22815, for 16- and 17-year-olds with a parent's consent); it limits the device (no more than "2.5 ounces net weight of aerosol spray"); and it makes using a spray "except in self-defense" an offence.⁶
• New York, Penal Law §265.20(a)(14) and (15), allows possession "for the protection of a person or property and use ... under circumstances which would justify the use of physical force pursuant to article thirty-five". Article 35 is the force law, §35.15 among it, that lesson 13 walked through. The exemption doesn't apply to anyone under 18 or with a felony or assault conviction, and sprays may be sold only by a licensed firearms dealer, "a pharmacist", or a vendor the state police approve.⁷
• Michigan bans possessing any device "designed to render a person temporarily or permanently disabled by the ejection, release, or emission of a gas or other substance", in MCL 750.224, and exempts "A self-defense spray or foam device as defined in section 224d".⁸ Section 224d sets that device's limits, makes using one on a person an offence, and then exempts, for the pepper (oleoresin capsicum) kinds it names, "reasonable use ... in the protection of a person or property under circumstances that would justify the person's use of physical force".⁸

Michigan's strength figure and the other device rules are left out on purpose: listed together, they would read as a buying guide. So is the federal rule on sprays in luggage when you fly, which is about travel rather than carrying where you live.

Put the three side by side and a pattern shows. It is this course's summary of three statutes, not a rule for all fifty: a spray of the kind each statute describes is lawful for many adults to have, some people are barred, and when one is used, the use is judged by the state's ordinary rules on force. That last point is lesson 13 again. A spray does not come with its own right to use it. And this course researched no figure on how well sprays stop attacks, so it gives none.

Bolton and Kalamazoo

This is general information, not legal advice. The law differs by state, and between England and Wales, Scotland and Northern Ireland. What follows asks which questions a court would ask. It gives no verdict.

Declan lives in Bolton. Twice in a month a man he doesn't know has followed him home from the bus stop, and the second time the man stood across the road for a few minutes before walking off. Declan is thinking about keeping a heavy torch in his coat, "just in case". On a call, his cousin Ade, who lives in Kalamazoo, Michigan, says she keeps a spray in her bag for exactly that reason.

Lesson 6 set out guidance, practitioner opinion in this course's labels, for the moment you think you're being followed. Being followed home more than once can be stalking, whoever is doing it, and the lesson on domestic abuse and stalking has the stalking lines.

In England and Wales, the questions about the torch are the 1953 Act's:

• Is it an offensive weapon? It was not made or adapted to injure. So the question is its purpose: does Declan have it with him intending to use it on a person? A torch carried "just in case" he's attacked, to hit with, has that purpose. A torch carried to see his way on an unlit path does not.¹
• Is he in a public place with it? The street and the bus stop are.¹
• Can he show a reasonable excuse? The burden is his. A continuing fear of the man who followed him, however genuine, is the kind of "constant or enduring threat" the CPS says is not enough; the exception is an attack that is imminent and specific.¹³

In Michigan, the questions about Ade's spray are different, and they come later:

• Is her device the kind section 224d describes? Section 224 bans devices designed to disable with a gas or other substance, and exempts that kind, so this is the question about having it, and the section sets its own limits.⁸
• If she ever used it: was it "reasonable use" to protect a person, in circumstances that would justify physical force under Michigan's force law?⁸ This course did not read Michigan's general self-defence law, so it cannot set out those circumstances.

Same fear, same intention, and two different legal questions. In England and Wales the intention is what can make an ordinary object an offensive weapon, before anything happens. In Michigan the sections this course read ask nothing about why Ade carries the spray: they ask what the device is, and the intention matters only if the spray is ever used. That difference is each legislature's choice, and the course calls neither better.

One line on lesson 6's options: the personal alarm there, practitioner opinion in this course's labels, is for setting off and leaving, and it is not a weapon under any rule read here.

Guns for protection: what the research argues

This section describes research, almost all of it American, because that's where the argument and the data are. It isn't guidance on guns, and in this course's labels every question in it is contested: the evidence points more than one way, and serious researchers disagree.

The broadest statement this course found comes from a 2013 report by the Institute of Medicine, working with the National Research Council, both part of the US National Academies, written to set priorities for research on gun violence. It holds both halves in one sentence: "Defensive use of guns by crime victims is a common occurrence, although the exact number remains disputed".⁹ The rest of this section fills in each half.

How often are guns used in self-defence?

The 2.5 million figure the report cites comes from a 1995 paper by Kleck and Gertz, who ran a national telephone survey of 4,977 people, reached by dialling numbers at random.¹⁰ Their estimate, in their words: "each year in the U.S. there are about 2.2 to 2.5 million DGUs of all types by civilians against humans, with about 1.5 to 1.9 million of the incidents involving use of handguns" (DGU is their abbreviation for defensive gun use). Their conclusion: "defensive gun use is very common in the U.S., and that it probably is substantially more common than criminal gun use."¹⁰ They also argued that the lower, government figure shouldn't be trusted: "There is little legitimate reason to continue accepting the NCVS estimates of DGU frequency as even approximately valid."¹⁰ The NCVS is the National Crime Victimization Survey, the US victim survey lesson 2 introduced.

A direct reply came from Hemenway in 1997, in the same journal. His central claim is about rare events: surveys "can wildly overestimate rare events which have some social desirability".¹¹ Social desirability is his term for an answer that makes the person giving it look good, which he also calls "personal presentation bias": someone who says he used a gun to ward off a criminal "is displaying the wisdom of his precautions and his capability in protecting himself".¹¹ If only a small share of people truly used a gun defensively, a few wrong "yes" answers swamp the true ones, and on one of his assumptions, 0.04% of adults, "the K-G figure of 2.5 million uses would be a thirty-three-fold overestimate!"¹¹

He then checks the high estimate against the victim survey's burglary figures. Kleck and Gertz's figures imply guns used defensively in about 845,000 burglaries in 1992. The victim survey counted fewer than 6 million burglaries that year, with someone certainly at home in 22% of them, about 1.3 million; fewer than half of US households had a gun, and in two-thirds of the occupied homes the people inside were asleep. So, he writes, the high estimate "asks us to believe that burglary victims in gun owning households use their guns in self-defense more than 100% of the time, even though most were initially asleep."¹¹

Kleck and Gertz answered in the same issue. Their central point is that the government survey undercounts because of how it asks: the NCVS "never directly asks about DGU. Instead it merely provides Rs with an opportunity to volunteer information about a DGU in response to a general question about self-protection actions" (Rs are respondents).¹² They answered the burglary check too. It treats the victim survey's count of burglaries as complete, they argue, when no criminologist believes the NCVS catches every crime. Defensive uses, they say, typically involve the defender breaking a law as well, such as having the gun unlawfully, so many never reach the victim survey at all. And the burglary figure rests on about 40 of their 194 reported incidents, a far smaller sample than their total.¹²

The National Academies report sets out the spread and gives a reason to doubt each end. Almost all national surveys, it says, put defensive uses at least as common as criminal uses, with "estimates of annual uses ranging from about 500,000 to more than 3 million", while "some scholars point to a radically lower estimate of only 108,000 annual defensive uses based on the National Crime Victimization Survey".⁹ The top figure, it says, "is based on an extrapolation from a small number of responses taken from more than 19 national surveys"; the low one "is difficult to interpret because respondents were not asked specifically about defensive gun use."⁹ (These are further published figures from each kind of survey, close to the 60,000 to 120,000 and the 2.5 million above.) It calls the variation "a controversy in the field". RAND's review of defensive gun use, updated in January 2026, puts the cause the same way: different definitions of defensive gun use and different ways of collecting information about it produce the wide differences.¹³

Why two surveys can be twenty times apart

This is the mechanism both sides are pointing at, with made-up numbers. The numbers below are this course's illustration, not anyone's finding; the two effects they show are the ones Hemenway and Kleck and Gertz each describe.

Imagine 10,000 adults answer a survey, and suppose that in truth 20 of them used a gun defensively last year, 0.2%.

Asked directly, with a few false positives. Suppose every one of the 20 says yes. Suppose too that 1% of the other 9,980 also say yes when they shouldn't, through misremembering, misreading the question or, on Hemenway's argument, liking the answer: that is about 100 people. The survey now counts about 120 yeses. Scale that up and the estimate is six times the truth, and five of every six yeses are wrong. The mistake rate was only 1%. It dominates because the true event is rare, which is Hemenway's argument, and the same arithmetic as Logic and Argument lesson 6's base rates.

Never asked directly. Now keep the same 20 true users, but ask them only a general question about anything they did to protect themselves, and suppose just 1 in 4 of them mentions the gun. Work out what this survey counts, and how far off it is, before you open the answer.

This is lesson 2's first question about any crime figure: who counted, and how.

Two more results belong here, one from each direction. On whether a defensive gun use protects the person using it, the National Academies report says studies "have found consistently lower injury rates among gun-using crime victims compared with victims who used other self-protective strategies", and that some "have found less loss and injury when a firearm is used".⁹ It adds that effectiveness is likely to vary with the victim, the offender and the circumstances, so the earlier findings need confirming or discounting.⁹ A 2015 analysis by Hemenway and Solnick, of the NCVS for 2007 to 2011, found that of over 14,000 incidents with the victim present, "127 (0.9%) involved a SDGU" (self-defence gun use), and that "After any protective action, 4.2% of victims were injured; after SDGU, 4.1% of victims were injured."¹⁴ In property crimes, 38.5% of victims who used a gun lost property, against 55.9% of victims who took protective action and 34.9% of those who used a weapon other than a gun.¹⁴ Their conclusion: "Compared to other protective actions, the National Crime Victimization Surveys provide little evidence that SDGU is uniquely beneficial in reducing the likelihood of injury or property loss."¹⁴ This course read the abstract only. It uses the survey Kleck and Gertz distrust, which is this course's observation, not the authors'.

What would settle it

It would take a survey that asks directly about defensive gun use, unlike the NCVS, and is built to measure false positives on a rare event, as Hemenway requires. It would also need checking against records of actual incidents. That is this course's reading of what each side's objection demands, and nobody has yet run such a study to both sides' satisfaction. Until then, "defensive gun use is rare" and "defensive gun use is common" are each a position in a dispute, not a finding.

Does a gun in the home change a household's risk?

The largest pooled study this course read is a 2014 meta-analysis by Anglemyer, Horvath and Rutherford in the Annals of Internal Medicine. It combined 16 observational studies comparing people with and without access to a gun in the home, and found pooled odds ratios of "3.24 (95% CI, 2.41 to 4.40)" for suicide and "2.00 (CI, 1.56 to 3.02)" for homicide victimisation, being killed.¹⁵ (Those are the corrected figures after an erratum.) The authors' conclusion is worded as an association: "Access to firearms is associated with risk for completed suicide and being the victim of homicide."¹⁵ And they name their limits: access was "determined by survey interviews in most studies; misclassification of accessibility may have occurred", and the pooled studies covered populations with different underlying risks.¹⁵ This course read the abstract only.

Two things about reading those numbers. First Aid and CPR lesson 4 showed that an odds ratio compares odds, not percentages, so 3.24 is not a 3.24% risk. When the outcome is rare, as these deaths are, odds and risk sit close together, so "about three times as likely" is a fair everyday reading, for the groups the studies compared rather than for any one household. What the number cannot tell you is why the groups differ: every study pooled was observational, so the finding is that access and death went together, not that one caused the other.

The other side of this question has a case of its own. Critics of these studies have published on how they are done. This course did not read that work, so it does not set out their argument, and that gap is the course's, not theirs. What it can give from what it read, though the first two items are about defensive use rather than household risk:

• The National Academies report's lower injury rates for gun-using victims, above.⁹
• Kleck and Gertz's point that most defensive uses in their survey involved no shot: only "24% claim to have fired their gun".¹⁰
• The report's own framing, which states the risk side as a condition to be tested: "if gun ownership raises the risk of suicide, homicide, or the use of weapons by those who invade the homes of gun owners, this could cancel or outweigh the beneficial effects" of defensive use. It adds that the early studies on this were not conclusive and the question needs careful research.⁹

An abuser's gun, and a victim's own

Lesson 7, "Domestic abuse and stalking", taught Campbell and colleagues' 2003 study of women killed by partners, compared with abused women who were not killed. Two of its findings belong here.

The abuser's gun. Access to a gun was among the factors that stood up when the others were taken into account: an adjusted odds ratio of 5.44 in the paper's fifth model (each model adds more factors, as lesson 7 showed).¹⁶ In a later model, once the abuser's use of a gun in the worst incident of abuse was added, access alone was no longer significant; the authors say the use apparently carried the effect of the access.¹⁶ That is the clearest finding in this section, and it points one way.

The victim's own gun. Here the study is much thinner, and it says so. Among women living apart from the abuser, sole access to a gun went with a lower risk in the model where it first appeared, and stopped being statistically significant once relationship factors were added.¹⁶ The authors' words, where the "case" women are those who were killed and the "control" women the abused women who weren't:

A victim's access to a gun could plausibly reduce her risk of being killed, at least if she does not live with the abuser. A small percentage (5%) of both case and control women lived apart from the abuser and owned a gun, however, and there was no clear evidence of protective effects.¹⁶

"No clear evidence" is not evidence of no effect, in either direction. It is a small group in one urban US study.

Advocates for survivors disagree about a victim's own gun, and each position here is quoted as a position. The National Domestic Violence Hotline, a US service, gives its position on its page about guns. Its reasons are the dangers it sees in a gun in a home where there is abuse, above all one an abusive partner can reach: "These increased risks are some of the reasons we do not advocate for survivors of domestic violence to purchase a gun." It also says to the person reading, "We at The Hotline know that you are the expert in your situation".¹⁷ America's 1st Freedom, a journal of the National Rifle Association, a gun-rights group, profiles Allyson Hottinger, a survivor and advocate who "urges fellow survivors to get their concealed-carry permit if comfortable". Her reason, in her words: the decision to carry "wasn't because I had in mind to kill my abuser, but it was because I had made the decision that if it came down to him or me, I was no longer going to be a victim of his abuse." And her qualification: "It might not be an option for everyone, but for the majority of survivors, it is."²¹

At the general level, the two sides of the US argument put it in their own terms. The NRA's lobbying arm, NRA-ILA, on its right-to-carry page: "Self-defense is a fundamental right", and "while law enforcement personnel provide protection generally, they are not required to protect individuals."¹⁸ Giffords, the gun-control group, states on its statistics page that "Access to a gun triples a person's risk of suicide", citing research this course did not read.²⁰

Each of those is a position, quoted as one. The evidence under them is the contested evidence above, and the decision it bears on, whether a particular person should keep a gun, is a value question with an empirical core. This course describes it and does not answer it.

What people get wrong

Each of the legal points is general information, not legal advice. The law differs by state, and between England and Wales, Scotland and Northern Ireland.

"Pepper spray is legal in the UK if it's for self-defence." In England, Wales and Scotland it's a prohibited weapon, and possession, even at home, is the offence. Purpose does not change that.⁵

"Carrying something ordinary for protection is fine, because it isn't a weapon." In England and Wales and in Scotland, the intention to use it on a person is what can make it an offensive weapon. In England and Wales a general fear, however genuine, is not a reasonable excuse under the CPS's guidance; how Scotland treats fear, this course did not read.¹³⁴

"A small folding knife is always legal to carry in England and Wales." One with a cutting edge of 3 inches or less is outside section 139, unless it locks, and a flick knife or gravity knife is an offence to have with you whatever its length. None of them is outside the 1953 Act if it is carried to use on someone.¹²³

"In the US, pepper spray is legal for anyone, anywhere." The three states read here each bar some people, regulate the device, or tie use to the force rules, and this course did not read the rest.⁶⁷⁸

"Defensive gun use is rare" and "defensive gun use is common, and that's settled." Each is one side of a dispute over estimates more than twenty times apart. The National Academies report calls it common, with the exact number disputed, and "common" is itself one side of that dispute, as "What would settle it" says.⁹

"The studies prove a gun in the home causes deaths" and "the studies prove nothing." They're observational and show an association, which the authors say. Critics have published on these studies' methods; this course did not read that work.¹⁵

Practice

Connections

Back. Lesson 13, "The law on using force", gave the questions a court asks about force actually used; this lesson adds the separate question of carrying, and the US spray statutes send you straight back to it. Lesson 14, "The home and property", took the same force law into the home. Lesson 6, "Out and away", asked about sprays abroad and kept the personal alarm for setting off and leaving. Lesson 7, "Domestic abuse and stalking", taught Campbell's study and the lines. Lesson 2, "Reading a crime number", taught that a figure depends on who counted and how, which is the whole of the defensive gun use dispute, and Logic and Argument lesson 6's base rates explain why a rare event is hard to count. First Aid and CPR lesson 4 explained odds ratios.

Forward. Lesson 16, "Afterwards", covers reporting, support and what happens after force is used. The carrying offences are one small corner of criminal law; the rest belongs to Introduction to Law, later on the Core.

Go deeper

• Institute of Medicine and National Research Council, Priorities for Research to Reduce the Threat of Firearm-Related Violence (2013), free to read. This course read chapters 2 and 4. The broadest short statement of what was known, and not known, about defensive gun use and household risk.
• RAND, "How Gun Policies Affect Defensive Gun Use", updated January 2026. This course read the summary and background. Useful for why the definitions differ.
• Kleck and Gertz's 1995 paper and Hemenway's 1997 reply, read in full for this course. Read them together; each is the other's best test.
• CPS, "Knife and Other Weapons Offences", updated May 2026, read in full for this course. How prosecutors in England and Wales read the carrying offences.

Sources

1. Prevention of Crime Act 1953, section 1, legislation.gov.uk, England and Wales. Read: in full; re-read on the live page, 25 September 2026.
2. Criminal Justice Act 1988, section 139, legislation.gov.uk, England and Wales version, with a separate Northern Ireland version on the same page. Read: in full; re-read on the live page, 25 September 2026.
3. Crown Prosecution Service, "Knife and Other Weapons Offences", prosecution guidance for England and Wales, updated 28 May 2026. Read: in full (7,300 words); re-read on the live page, 25 September 2026.
4. Criminal Law (Consolidation) (Scotland) Act 1995, section 47 and section 49, legislation.gov.uk. Read: the sections; re-read on the live pages, 25 September 2026.
5. Firearms Act 1968, section 5, legislation.gov.uk, extent England, Wales and Scotland. Read: in full; re-read on the live page, 25 September 2026. The penalty provisions were not read.
6. California Penal Code, §22810, with §22815 and the definition in §17240. Read: the sections; re-read on the live pages, 25 September 2026.
7. New York Penal Law, §265.20(a)(14) and (15). Read: (14) and (15) in full; re-read on the live page (browser), 25 September 2026.
8. Michigan Compiled Laws, §750.224 and §750.224d. Read: the sections; re-read on the live pages, 25 September 2026.
9. Institute of Medicine and National Research Council, Priorities for Research to Reduce the Threat of Firearm-Related Violence (National Academies Press, 2013). Read: chapters 2 and 4; re-read, 25 September 2026.
10. Kleck, G. and Gertz, M., "Armed Resistance to Crime: The Prevalence and Nature of Self-Defense with a Gun", Journal of Criminal Law and Criminology 86(1), 1995, pp. 150 to 187. Read: in full, as the Internet Archive saved it (capture 20260831195132).
11. Hemenway, D., "Survey Research and Self-Defense Gun Use: An Explanation of Extreme Overestimates", Journal of Criminal Law and Criminology 87(4), 1997. Read: in full, as the Internet Archive saved it (capture 20260919163504).
12. Kleck, G. and Gertz, M., "The Illegitimacy of One-Sided Speculation: Getting the Defensive Gun Use Estimate Down", Journal of Criminal Law and Criminology 87(4), 1997. Read: in full, as the Internet Archive saved it (capture 20240709092114). Its argument is used; its opening remarks about its critic are not.
13. RAND, Gun Policy in America, "How Gun Policies Affect Defensive Gun Use", updated January 29, 2026. Read: the summary and background; re-read, 25 September 2026.
14. Hemenway, D. and Solnick, S. J., "The epidemiology of self-defense gun use: evidence from the National Crime Victimization Surveys 2007-2011", Preventive Medicine 79, 2015, pp. 22 to 27. Read: abstract only.
15. Anglemyer, A., Horvath, T. and Rutherford, G., "The accessibility of firearms and risk for suicide and homicide victimization among household members", Annals of Internal Medicine 160(2), 2014, pp. 101 to 110, with its erratum. Read: abstract only.
16. Campbell, J. C. and colleagues, "Risk factors for femicide in abusive relationships: results from a multisite case control study", American Journal of Public Health 93(7), 2003, pp. 1089 to 1097. Read: full text (PMC); re-read, 25 September 2026. The 5.44 is from the paper's fifth model.
17. National Domestic Violence Hotline, "Safety Planning Around Guns and Firearms". Read: full; re-read, 25 September 2026. An advocacy and service page, cited for its position only.
18. NRA Institute for Legislative Action, "Right-to-Carry and Concealed Carry". Read: the lines quoted; re-read on the live page, 25 September 2026. An advocacy page, cited for its position; its statistics are not used.
19. Support lines in the content note, each read on its service's own pages on 25 September 2026: Victim Support (England and Wales), get help and about us; Victim Support Scotland; Victim Support NI, home page and contact; and VictimConnect, home page and get help. Read: each page in full, re-read 25 September 2026.
20. Giffords Law Center, "Gun Violence Statistics". Read: the page, 25 September 2026. An advocacy page, cited for its position; the research it cites was not read.
21. Susanne Edward, "Self-Defense: A Women's Movement", America's 1st Freedom (a journal of the National Rifle Association), 27 April 2021. Read: in full, 25 September 2026. An advocacy page, cited for the position of the survivor it quotes; nothing else in it is used.
22. GOV.UK, "Guns, knives, swords and other offensive weapons: UK border control", updated 27 February 2025. Read: in full, 25 September 2026.