The historical claim of this section is that Graunt, in 1662, did not take the London bills of mortality as given. He described how they were made, judged the searchers' reports cause by cause, accepted the coarse categories wherever his purposes allowed, and corrected the count of plague deaths by setting it against ordinary years and against the christenings. He addressed the result, explicitly, to men in government. For the debate the section adds this: the definitions in Graunt's text are stated as open rules that any reader may check, which fits Porter's account, yet they are fixed for the uses of magistrates and councillors, which fits Desrosières's, and they rest on the knowledge of the very women whose judgments Graunt corrects, which Scott's opposition of state and local knowledge does not anticipate.
4.1From a death to a printed sheet
Graunt's account of how the bills were made is brief and exact. A death first becomes known by sound or by request: "When any one dies, then, either by tolling, or ringing of a Bell, or by beſpeaking of a Grave of the Sexton, the ſame is known to the Searchers, correſponding with the ſaid Sexton."
The searchers then go to look. "The Searchers hereupon (who are antient Matrons, ſworn to their Office) repair to the place, where the dead Corps lies, and by view of the ſame, and by other enquiries, they examine by what Diſeaſe, or Caſualty the Corps died." Two sources of knowledge are named in that sentence: what the searchers see, and what they learn by asking. Neither is a physician's examination.
The report then moves upward and outward on a weekly timetable. "Hereupon they make their Report to the Pariſh-Clerk, and he, every Tueſday night, carries in an Accompt of all the Burials, and Chriſtnings, hapning that Week, to the Clerk of the Hall. On Wedneſday the general Accompt is made up, and Printed, and on Thurſdays publiſhed, and diſperſed to the ſeveral Families, who will pay four ſhillings per Annum for them."
The chain runs from a bell or a request at the churchyard, through the searchers' view, the parish clerk and the Clerk of the Hall, to the printer. Its last link points outward, to subscribing families, not only upward to authority. Graunt adds a memorandum that the original entries were, if anything, fuller than what was printed: "the Original Entries in the Hallbooks were as exact in the very firſt Year as to all particulars, as now; and the ſpecifying of Caſualties and Diſeaſes, was probably more."
Set against Foucault's plague town, the chain is recognisably a register of every death passed through officers to a centre (Foucault 1977), and Slack's account places searchers and bills among the measures of the plague orders (Slack 1985). Graunt's description differs in one respect that bears on the debate. The register does not stop at the centre. It is printed and sold. The knowledge it holds is offered, for a fee, to anyone in the city who wants it.
4.2For whom the count was kept
Graunt was puzzled by that publicity, and said so. The opening of his chapter on the casualties states the obvious use of the bills for those in charge: "There ſeems to be good reaſon, why the Magiſtrate ſhould himſelf take notice of the numbers of Burials, and Chriſtnings, viz. to ſee, whether the City increaſe or decreaſe in people; whether it increaſe proportionably with the reſt of the Nation; whether it be grown big enough, or too big, &c." He then doubts the other use: "But why the ſame ſhould be made known to the People, otherwiſe then to pleaſe them as with a curioſity, I ſee not."
He found the same incuriosity among the buyers. His preface begins with the readers of the weekly sheets, who "made little other uſe of them, then to look at the foot, how the Burials increaſed, or decreaſed", and who, in time of plague, read them "that ſo the Rich might judge of the neceſsity of their removall, and Trades-men might conjecture what doings they were like to have in their reſpective dealings". The public use he describes is private and short-term: whether to leave town, how trade will go. The use he proposes is public and long-term, and he directs it upward.
The dedication makes the address explicit. Graunt writes to the Lord Privy Seal that he has "reduced ſeveral great confuſed Volumes into a few perſpicuous Tables", and hopes the result "may be of as much uſe to Perſons in your Lordſhip's place, as they are of little or none to me". He then lists the positions he thinks a peer or councillor should consider, among them "That the troubleſome ſecluſions in the Plague-time is not a remedy to be purchaſed at vaſt inconveniencies", "That the greateſt Plagues of the City are equally, and quickly repaired from the Country", and "That the number of Heads is ſuch, as hath certainly much deceived ſome of our Senatours in their appointments of Polemoney". The bills, in this framing, are an instrument for judging the plague orders themselves, for setting a poll tax, and for estimating the city's strength. The dedication closes by inviting the Lord Privy Seal and the world to "ſee the Wiſdom of our City, in appointing, and keeping theſe Accompts".
This address fits Desrosières's account of statistics bound up with the making of states. The first of Graunt's listed positions also connects to Slack's account of magistrates divided over whether isolation was just (Slack 1985). Graunt offers his tables as evidence in that dispute, on the side of doubting that shutting up houses was worth its cost. The dedication does not say how the tables show this, and the chapters read here do not argue it directly.
4.3A count offered for correction
The same preface states the opposite principle, in terms close to Porter's. Graunt describes collecting "as much matter of that kind, even as the Hall of the Pariſh-Clerks could afford me", reducing it to tables "ſo as to have a view of the whole together", and then testing his first impressions against them. His defence of the work, as one "not profeſſing Letters", is that he has made his tables public so that others can check him: "I have taken the pains, and been at the charge, of ſetting out thoſe Tables, whereby all men may both correct my Poſitions, and raiſe others of their own: For herein I have, like a ſilly Schole-boy, coming to ſay my Leſſon to the World (that Peeviſh, and Tetchie Maſter) brought a bundle of Rods wherewith to be whipt, for every miſtake I have committed."
Here the count is meant to travel. The author disclaims learned standing and offers his materials in place of it. Porter's argument that rules are pursued most by weak or contested experts finds a plain instance: Graunt presents himself as an outsider to the learned, and asks to be trusted for the tables rather than for himself. Rusnock's stress on gathering is met too, since the preface says where the numbers came from (the Hall of the Parish Clerks) and the first chapter says how they were made. The count is thus offered upward, to a Lord Privy Seal, and outward, to "all men", and the document is built to do both.
4.4What the searchers could be trusted to see
The core of the chapter on the casualties is Graunt's judgment of the searchers. He begins from the use he wants the bills to serve, which is to show "That the ſtate of health in the City may at all times appear." He answers the objection that only the plague matters with the sentence the rules allow to be quoted in full: "That the knowledg even of the numbers, which die of the Plague, is not ſufficiently deduced from the meer Report of the Searchers, which onely the Bills afford; but from other Ratiocinations, and comparings of the Plague with ſome other Caſualties." He promises a correction: "For we ſhall make it probable, that in Years of Plague a quarter part more dies of that Diſeaſe then are ſet down; the ſame we ſhall alſo prove by the other Caſualties."
Before correcting the searchers, he asks how far they can be believed: "Now, to make theſe Corrections upon the perhaps, ignorant, and careleſs Searchers Reports, I conſidered firſt of what Authority they were in themſelves, that is, whether any credit at all were to be given to their Diſtinguiſhments". His answer is a rule of sorting. Some categories are "but matter of ſenſe": whether a child was abortive or stillborn, whether a person was aged. For those, "I concluded the Searchers Report might be ſufficient in the Caſe."
The definition of the aged is stated twice, and not identically. In this paragraph the searchers' aged are those "above ſixty years old, or thereabouts", judged "without any curious determination" of whether such persons died "purely of Age". Later in the same chapter, when Graunt uses the aged as his standard of a healthy place, he writes: "Onely the queſtion is, what number of Years the Searchers call Aged, which I conceive muſt be the ſame, that David calls ſo, viz. 70." The two passages do not contradict each other outright, since the first describes what the searchers can see and the second what Graunt supposes they mean. But they show the definition being fixed by Graunt's own conjecture at the point where he needs it, and the second passage frames it as a question.
For other categories Graunt lowers the standard deliberately, by appeal to purpose. Of consumption: "As for Conſumptions, if the Searchers do but truly Report (as they may) whether the dead Corps were very lean, and worn away, it matters not to many of our purpoſes, whether the Diſeaſe were exactly the ſame, as Phyſicians define it in their Books." Of a man who dies old of a cough, Graunt says that it is "little errour (as to many of our purpoſes)" if he is counted among the aged. Of sudden death: "Again, if one died ſuddenly, the matter is not great, whether it be reported in the Bills, Suddenly, Apoplexie, or Planetſtrucken, &c."
The infants are the clearest case of a definition he could not settle. "In the matter of Infants I would deſire but to know clearly, what the Searchers mean by Infants, as whether Children that cannot ſpeak, as the word Infans ſeems to ſignifie, or Children under two or three years old". He does not need to know what each infant died of. He needs to know where the line falls, because his purpose is age: "for, I ſay, it is ſomewhat, to know how many die uſually before they can ſpeak, or how many live paſt any aſsigned number of years."
His general rule follows: "I ſay, it is enough, if we know from the Searchers but the moſt predominant Symptomes". And he adds that the searchers are not alone in their knowledge: "In many of theſe caſes the Searchers are able to report the Opinion of the Phyſician, who was with the Patient, as they receive the ſame from the Friends of the Defunct". Where the casualties are drowning, scalding or smallpox, the searchers' "own ſenſes are ſufficient". The doubtful ground is marked off for separate treatment: "thoſe Caſualties, which are apteſt to be confounded, and miſtaken".
Read against Scott, this passage runs the other way from an imposition of simple categories over local knowledge. The simple categories already exist; they are the searchers' own. Graunt accepts them where their coarseness does not matter to his question, distrusts them where it does, and builds his corrections on comparisons among the numbers the searchers produced. The local knowledge, the matrons' view and enquiries and what the friends of the dead told them, is not overridden. It is the raw material, judged category by category. Hacking's point that categories shape what is counted is visible in the infants: what Graunt can learn about early death depends on a word whose limits he cannot discover from the bills. Scott's argument concerns schemes that imposed simplification and failed; Graunt imposed nothing, he read, and his text cannot show whether later uses of the bills overrode local knowledge. It shows a reader of official counts treating their categories as conventions of varying reliability, and saying which.
4.5Proportions as a standard
Graunt's next step is to turn counts into proportions, and proportions into standards for judging the city. The first observation in the chapter sets a group of children's causes against the total of burials "in twenty Years" and concludes "that about ⅓. of the whole died of thoſe Diſeaſes, which we gueſs did all light upon Children under four or five Years old." The word "gueſs" is his; the age limit is inferred from the names of the causes, not reported.
He then offers the large table of notorious diseases, so "that the reſpective numbers, being compared with the Total 229250, thoſe perſons may the better underſtand the hazard they are in." The total is the denominator, and the purpose is stated: the reader who fears a particular disease can set its number against all burials. This is a denominator chosen for the frightened private reader, not for the magistrate, and it sits in the same chapter as the defence of the bills as instruments of state.
The chapter ends with the observation the reading note on Graunt records as one of his regularities: "among the ſeveral Caſualties ſome bear a conſtant proportion unto the whole number of Burials". The chronic diseases and even some accidents keep that proportion, while the epidemic diseases do not, "ſo as in ſome Years, or Moneths, there died ten times as many as in others." Graunt does not call the constancy a law. He uses it to separate two kinds of cause: those whose share of deaths is steady and those that come in surges. That separation is what makes his correction of the plague count possible.
Hacking's account places the idea of statistical laws of society in the nineteenth century, after the avalanche of printed numbers (Hacking 1990). Graunt's constant proportions, drawn in 1662 from printed weekly sheets, are an earlier observation of steadiness in counts of death. The text does not claim more than steadiness; it offers no theory of why the proportions hold.
4.6Correcting the plague
The chapter on the plague begins with the great mortalities of the age and their figures, as given in the bills, and then states the problem that raw totals cannot settle: "Now it is manifeſt of it ſelf, in which of theſe years moſt died; but in which of them was the greateſt Mortality of all Diſeaſes in general, or of the Plague in particular, we diſcover thus." The answer comes from two proportions. The first is plague deaths to all deaths. The second is christenings to burials, which stands in for the size of the living population.
By the first proportion, 1603 was the worst plague year. By the second, 1603 and 1625 were equally the worst years of mortality. Graunt states the conflict openly: "We ſaid that the year 1603. was the greateſt Plague year. And now we ſay, that the ſame was not a greater year of Mortality then Anno 1625." He resolves it by doubting the classification of causes in 1625, saying that "more died of the Plague then were accompted for under that name."
The proof is a comparison with ordinary years: "In the ſaid year 1625. there are ſaid to have died of the Plague 35417. and of all other Diſeaſes 18848. whereas in the years, both before and after the ſame, the ordinary number of Burials was between 7. and 8000." The deaths from other causes in 1625 far exceed what ordinary years would lead one to expect. Graunt attributes the excess to the plague, which brings the share of plague deaths in 1625 into line with 1603. He then generalises: "From whence we may probably ſuſpect that about ¼. part more died of the Plague then are returned for ſuch". The 1636 bills are offered as a second test of the same quarter.
Three features of this argument bear on the question of what had to be fixed before a count could be used. First, the denominator is not the population of London, which no one had counted. It is the ordinary number of burials in neighbouring years, taken as the expected level, and, for the second measure, the christenings. Second, the correction is a correction of definition. Graunt does not say more people died than were buried. He says some deaths were placed under the wrong name. Third, the claim is hedged: "probably ſuſpect", and earlier "make it probable". The reading note records that he observed plague years could be told from the excess of burials over ordinary years (Graunt 1662). The text shows that the same excess was also his tool for correcting the searchers' own attribution of cause.
In the terms of Lipsitch and colleagues, the distortion Graunt corrects is one of case definition and reporting (Lipsitch and others 2015). The searchers, for reasons Graunt does not give, returned some plague deaths under other names. His baseline is itself a count by the same searchers: he corrects them with their own numbers.
The same chapter draws a conclusion about cause that the counts cannot fully carry. From the long persistence of plague after 1636 and its sudden weekly jumps, Graunt infers "that the Contagion of the Plague depends more upon the Diſpoſition of the Air, then upon the Effluvia from the Bodies of Men." The bills record deaths by week. They cannot record the air. The inference is offered as a proof, and it is the point in the text where a count carries a causal claim most heavily.
4.7Christenings as the standard
Chapter V makes the christenings the explicit standard for the size of the living city. Its first sentence holds that the "Decreaſe, and Increaſe of People is to be reckoned chiefly by Chriſtnings, becauſe few bear children in London but Inhabitants, though others die there." The definition of the denominator is argued, not assumed. Burials include strangers who die in London; christenings do not. Graunt then admits a weakness and absorbs it. The accounts of christenings were well kept until differences in religion caused neglect, "although even theſe neglects we muſt confeſs to have been regular, and proportionable."
That clause is a small theory of error. A count with a steady shortfall can still serve as a standard, because comparisons between years are unaffected. On that basis Graunt reads the christenings, "which are our Standard in this Caſe", to show how quickly London refilled after a plague, and concludes that the new people came from outside: "Ergo, it muſt be by new Affluxes to London out of the Countrey." His conclusion then returns to an interest beyond public health: "let the Mortality be what it will, the City repairs its loſs of Inhabitants within two years, which Obſervation leſſens the Objection made againſt the value of houſes in London".
4.8The record beside the text
Skeptron's record is not a summary of Graunt's argument. It is a description of a copy and a warning about how to use it. Its Scenario opens with the edition:
The relevant copy at the Wellcome is the 1665 fourth edition printed at Oxford and includes 205 pages and folding tables, among other features.
It records the attribution:
A bibliographic note is attached stating that it has also been attributed to William Petty in addition to John Graunt's name.
Its third question asks whether the date of the copy may stand for the date of the data: "May the publication year of the copy be used as the observation year or first-edition year of every table?" The answer declines, and its reason is a list of things to keep apart:
The edition discrepancy, the attribution note, the period covered by the tables, the mortality classification, and the actual onset time are kept distinct.
Two items in that list correspond to things the text shows. One, "the mortality classification", is exactly what Graunt spends his chapter on the casualties judging. The other, "the period covered by the tables", matters because Graunt's figures run across decades of bills, from the first of "the years 1592, and 1593, 1603, 1625, and 1636" to his own time; a count printed in 1662, or in a copy of 1665, is not a count of 1662 or 1665. The record's caution about "the actual onset time" keeps the date of a printing apart from the date of the deaths it reports; a copy dated 1665 is not, by its date alone, evidence about that year's mortality.
The record differs from the text read here in one respect that should be stated plainly. The record describes the fourth edition of 1665; the text is the first edition of 1662. The record gives no figure from the bills and quotes nothing of Graunt's argument. It does not mention the searchers.
What the modern summary keeps is the separations: copy from first edition, publication from observation, classification from cause, attribution from name. What it leaves out is Graunt's own argument, in which one of those separations, of the searchers' classification from the cause of death, was already being made inside the book.
4.9What this case adds
Graunt's text gives a document-level answer to the article's question. Before his count could be used, he had to fix which of the searchers' categories could be trusted and for what purpose, what "Aged" and "Infants" meant, which baseline (ordinary years, or christenings) stood for the expected level, and how much of a misclassified cause to restore. He fixed each openly, in a form a reader could check against the tables he printed. For the debate, the case supports Porter's link between rules and weak or contested standing, and Desrosières's account of classifications at once technical and political, together in a single document. It sits awkwardly with Scott's opposition: the count is built from local knowledge and corrected by comparison, not imposed upon it.