Clocks and their visibility in churchwardens’ accounts

It is important to establish the reliability of churchwardens’ accounts as a source showing the presence or absence of church clocks in early modern England. Various questions arise, some with respect to each parish individually, and others to the overall pattern of accounts and clocks recorded therein. Questions of both positive evidence and negative evidence potentially arise for every individual parish. Does recorded spending on a clock in churchwardens’ accounts necessarily mean that there was a clock in the church? And does the absence of a clock from the accounts mean that there was no church clock? Other questions arise when attention shifts from single parishes to the overall pattern of surviving accounts, about both the apparent patterns of clocks, and the representativeness of the extant accounts. The various issues are considered here as four questions.

1). How likely are clocks to be evident in churchwardens’ accounts?

2). Are parishes with clocks more likely to have extant accounts?

3). Do patterns of account survival bias estimates of the numbers and geographical distribution of clocks?

4). Can details of clock mechanisms and signalling be discerned from accounts?

1). How likely are clocks to be evident in churchwardens’ accounts?

There arrangements through which church clocks were bought and maintained varied, and different funding arrangements were more or less conspicuous in churchwardens’ accounts. It is useful to sub-divide the question into the visibility of the initial acquisition of the clock, and the visibility of its subsequent maintenance.

The initial acquisition of a clock could take one of three forms, which impacted differently on the churchwardens and their accounts.

First, some clocks were gifts to parishes. The donor was usually an individual, but sometimes a family, or a sub-group of parishioners. Whether a donation is evident in the accounts depends first on whether it was made via the churchwardens rather than to their superiors, and second on whether the donation involved just the clock itself, or met the costs of its installation. Donors’ practices varied in this regard: some provided just the clock itself, others gifted substantial sums of money or rents from houses or land to cover associated costs, and such fully-funded clocks were the least likely to appear in the churchwardens’ accounts. They might, though, be recorded on a plaque on the clock, or alongside it in the clock-room.

Second, a clock could be bought from parish funds, in which case it should feature within the churchwardens’ expenditure account, or discharge.

However, the high cost of clocks meant that parish purchases were likely to entail a third form of funding, through the levying of a special rate for the purpose. This would appear on the income account, and where clock commissioning and construction extended over several months (as was usually the case), it was also likely to figure among the sums passed from one pair of wardens to the next. Occasionally, though, a sub-group of parishioners were responsible for assessing and collecting contributions, and negotiating with a clocksmith, and their accounts were entirely self-contained. Much depends here on whether they were kept alongside the churchwardens’ accounts and other parish records.

Turning to records of maintenance, keeping a church clock running involved significant ongoing costs. Clocks varied greatly in the quality of their construction and their smooth long-run operation, but it is essential to appreciate that all public clocks required much ongoing maintenance to keep running. Some were complex and potentially costly: replacing major parts of the mechanism, providing a new dial or hand(s), rebuilding a chime barrel. Others were the utterly normal ‘minding of the clock’: the daily winding, frequent oiling, replacement of worn minor parts. Spending went both on labour, the wages of the person (usually the sexton or the parish clerk) who wound and set the clock every day, and on materials, especially the oil or grease used to lubricate the clock’s moving parts, and occasional purchases of cords or rope.

Even for well-endowed clocks, it was rare for early modern donors to remove all these almost-constant maintenance responsibilities from the parish, and routine low-level spending routinely and frequently appears in accounts. Almost unique in keeping such spending completely out of churchwardens’ accounts were some pre-Reformation instances of parish gilds, whose whole existence hinged around their members funding the clock, and performing care and maintenance tasks. These few early cases apart, the flow of costs to churchwardens from maintenance work was sufficient for periods without it to be unusual, and suggestive that the clock had broken down. Even in the extreme case where a clock and its care were wholly funded by endowment, and met through special administrative arrangements, clocks, clock-bells, or clock-dials were liable to be mentioned by churchwardens in passing. For example, wardens might employ the same craftsmen or suppliers of materials, in paying for other work for other work while they were engaged in work on the clock. Or a clock could be mentioned in passing whilst describing other of their activities or responsibilities situated near the clock, or in or adjacent to the clock room.

In parishes that maintained a clock from parish resources overseen by the churchwardens, consequently, the critical factor is whether their accounts are sufficiently detailed to reveal the presence or clear absence of such spending. The key practical difficulties are three-fold. First, the audited accounts copied into parish account books were aggregate summaries of churchwardens’ spending, and much detail was compressed. Second, exactly the routine character of this spending made it among the most likely to be much abbreviated in both the original accounts, and the aggregate summary. And third, in the long-run, the degree of aggregation and compression becomes much greater, in all parts of England. Inevitably, the greater the compression, and the more that payments in the final accounts are primarily concerned to identify the recipient rather than record details of work and materials, the more uncertainty surrounds the work for which payment had been made.

The difficulties should not be exaggerated, though, especially before the later-seventeenth century. It is certainly true that the presence or absence of a clock will be a lot clearer than the exact details of clock and time-signals, but the recurrent nature of such spending enables considerable inferences to be made even where specific details of expenditure are only occasionally stated, so long as reasonable numbers of account survive. The most problematic instances are those combining highly aggregated accounts, on the one hand, and survival in very limited numbers, on the other. Mercifully, this combination is rare among the nearly 1200 parishes studied.

2). Are parishes with clocks more likely to have extant accounts?

The most straightforward statistical inference from churchwardens’ accounts would take those extant to provide a random sample of parishes. Yet it is clear that documented parishes are disproportionately likely to be in towns than the countryside, and that accounts from parishes with small populations are comparatively rare. These complicating effects on analysing documented clocks are compounded if the survival of accounts was itself linked to clock-possession. This is not to suppose some quasi-mystical association, simply that a clock could more feasibly be afforded by a parish of some wealth and a capacity to organise its affairs, and that churchwardens’ accounts were more likely to be kept, audited, and preserved in such a parish than one with the opposite characteristics. This seems plausible, even obvious.

That the likelihoods of parishes having a clock and churchwardens’ accounts probably co-vary needs to be borne in mind, but this is primarily a warning against simplistic generalisation about the numbers and geographies of clocks from surviving accounts, rather than an insuperable obstacle to exploration of them. However, it is necessary to scrutinise how the likely effects can be countered.

3). Do patterns of account survival bias estimates of the numbers and geographical distribution of clocks?

As noted above, certain patterns in the survival of churchwardens’ accounts bias any simple assessment of national patterns of church clocks. Not only do extant churchwardens’ accounts demonstrably come from an unrepresentative sample of English parishes, but the degree of unrepresentativeness changes over time.

This is clear both from the overall composition of documented parishes at any given time, between urban and rural parishes, for example, and among different parts of England, and among parishes within regional and settlement types. The biased survival of accounts comes from the overlaps among the sorts of parishes having clocks, the sorts of parishes keeping accounts, and the sorts of places preserving their accounts. All three seem associated with the general size, organisation, and social complexity of parishes.

Churchwardens’ accounts survive more widely, more systematically and from earlier dates, in populous and wealthy urban parishes. This is especially evident before the Reformation, with endowed supplementary provision in worship, building, liturgical and secular objects, and charity. It is highly likely that such places contained more parishioners familiar with writing, arithmetic and account-keeping than small, rural parishes. Besides greater likelihood of account-keeping, various 'survival-affecting' factors were similarly unevenly distributed. Notions of archived accounts as resources and precedents for later churchwardens, and a more general valuing of the idea of ’corporate' records’ are plausibly a sensibility primarily encountered in environments of denser institutional or corporate activity, though not an exclusively urban sensibility.

Greater sensibility to administration of parish activities could affect the preservation and survival of accounts, especially more secure storage locations and better storage conditions. For example, some parishes possessed a dedicated and lockable parish chest for storage, located in a dry, secure room in a church tower, while many others seem to have made do with keeping their old records in a more open and disturbed location.

Summarising: there are several plausible reasons why parishes whose churchwardens’ accounts survive were more likely to have had church clocks than the parishes whose accounts are not extant. The accounts database is hence likely to over-estimate proportions of parishes with clocks, because certain types of parishes are over-represented within it. It is therefore necessary first, not to be uncritical about overall national or regional trends from trends among the documented parishes, and second, to take opportunities to identify the scale and trends of this bias as the volume of surviving material increases over time.

4). Can details of clock mechanisms and signalling be discerned from accounts?

Details of the material and human facets of timekeeping, the performance of clocks, and the priorities of parishioners are all reliant on the degree of detail in the accounts. As noted, the presence or absence of a clock will be clearer than the details of mechanisms and signalling. Much depends on whether distinctive terminology was involved, as - despite much leeway in spelling - with mentions of ’pendiloe’ or ’pendolum’ in the late-seventeenth century, or ’carilon’ [chime] barrels a century earlier. Without distinct terminology, recording is often insufficiently detailed for precise identification of the clock mechanism. The nature of clock escapements is often unclear, as is whether a clock-dial had one or two hands. While the sums spent and personnel involved may be suggestive, but the evidence is not conclusive.

Clarifying the nature of a particular clock once again faces the twin obstacles of the brevity of descriptions and the generality of terms used, on the one hand, and limited detailing of objects purchased or work performed, on the other. These problems are exacerbated when extant accounts form an incomplete and discontinuous series. From a single reference it may be unclear whether spending on a ’dial’ is a clock-dial or a sundial, without some fleshing out of the circumstances of work, the scale of the object, the nature of related expenditure, and so on.

Overall, the visibility of clocks in accounts varied with their complexity, partly because better-performing clocks entailed specialised components and distinctive terms. From the earliest parish clocks to the technological novelties after 1660, several new technological elements involved some neologisms, but many clock components continued to be described very generally: ropes, a wheel, wire, an iron bar, a pin, oil. Often, such terms provide an insufficient basis from which to infer the object worked on.

Insufficiently specific references to clocks make some churchwardens’ accounts, like all other forms of contemporary documentation, ambiguous. The problem is unavoidable. Figures for explicitly identified features of clocks will therefore underestimate their frequency, and should be understood as minima. The scale of underestimation remains uncertain, however, and it is conceivable that major changes in it could create deceptive trends in the apparent presence of particular technological or design elements. That possibility should be considered case-by-case.

Conclusions

After the Reformation, churchwardens were central to routine parish and church administration, and a series of churchwardens’ accounts, provided that they are not too heavily compressed into aggregate summary form, provides reliable evidence as to whether parishes possessed church clocks. This applies to accounts as sources both of positive evidence of a clock’s presence, and of negative evidence that there was not a church clock. While there are various reasons why a continuously operating clock maintained by a parish clerk or sexton, and paid for by the churchwardens, might escape explicit mention in a single account, this would be an unusual occurrence anywhere that detailed churchwardens’ accounts survive in any numbers.

Similarly, it was theoretically possible for a clock to be present without the churchwardens ever spending money in relation to it, or mentioning it in passing, but it is hard to point to a single instance among the 1000-plus parishes studied here.

Complications inevitably attend parishes where the survival of accounts is restricted to infrequent single years or shorter periods of less than a year, and in gauging any short run of accounts, the level of disaggregation in detailed spending is critical.