Reviews
Part of 9 points on people by century that matter
People by century directory, checked against the record in 2027
A people by century directory needs one stated bucketing rule, honest treatment of boundary lives, and a scope note on what each period page misses.
A period index is a browsing structure, and browsing structures fail differently from datasets. The data can be right while the index quietly teaches the reader something false, usually that each person belongs in one box.
This page is about designing that index: how to bucket a life that spans a boundary, how to cross-reference without duplicating, and what to put on the bucket page itself.
What to take away
- A person with a long life belongs in more than one bucket, and the index should say so rather than choosing.
- Which date puts someone in a bucket is a rule, and it has to be visible to the reader, not just to you.
- Cross-references cost nothing and prevent the commonest reader error.
- A bucket page that does not say what it excludes will be read as complete.
Three bucketing rules, and what each teaches the reader
| Rule | Puts a person in | Reader takes away |
|---|---|---|
| Birth date | The period they were born into | A cohort view, weak for anyone who worked late |
| Peak or main activity | The period they are known for | An activity view, requires a judgment per person |
| Any overlap | Every period they lived through | A presence view, larger buckets, no judgment needed |
None is correct in general. What is not acceptable is mixing them silently, which is what most published period indexes do: entries arrive from different sources under different conventions and the index inherits all of them.
Pick one rule, state it on every bucket page in a sentence, and store the rule alongside each entry so a later switch is a re-derivation instead of a rebuild.
Boundaries deserve their own treatment
The interesting entries are the ones near a boundary, and they are exactly the ones a single-bucket index hides. Two habits fix most of it.
- Give each bucket page a short section for entries that straddle its start or end, and link them to the neighboring bucket in the same line.
- Let one entry appear in two buckets with a marker saying which is the primary, rather than duplicating the record. Duplicated records diverge.
The boundary section is also where date precision becomes visible. An entry recorded as a decade cannot be placed on the correct side of a boundary at all, and saying so in the index is more honest than picking a side.
What belongs on a bucket page
- The bucketing rule, in one sentence.
- The date range the bucket covers, including whether the endpoints are inclusive.
- A scope note: which sources were searched for this period and which were not.
- A note on coverage, because earlier periods have fewer surviving records rather than fewer people.
- Links to the neighboring buckets, at the top and not only at the bottom.
Point four matters more than it looks. Every period index makes recent periods look busier, which readers interpret as history rather than as survivorship bias in the record. One sentence on the page prevents that reading.
Naming buckets honestly
Calendar centuries are arbitrary, but at least everyone understands how. Named periods are worse in an index because they carry an argument, and different fields define them differently.
They do not apply outside the region where they were coined. Using them as browse labels imports an entire historiography into what the reader thinks is a neutral list.
If a named period is genuinely useful for your subject, use it as a secondary label under the century, define it on the page, and say whose definition you used.
What a period index will not sort by here
It does not sort or subdivide people by origin, ancestry, community, or nationality. A period bucket subdivided that way stops being a browse structure and becomes a tally, and the reasoning against that is on the heritage pillar and, for ranked forms of it, the fame rankings pillar.
For the dating arithmetic behind the buckets, see the century pillar; for how these records are found, the historical figures pillar; and for the jurisdiction fields the same entries carry, the country pillar.
Common questions
Should a person appear in more than one century page?
Yes, when they worked across a boundary, with one page marked primary. Hiding a working life inside one bucket is a bigger error than a second listing.
Is "early" and "late" within a century useful?
Only if the underlying dates support it. With decade-level precision it invents a distinction the sources do not carry.
How should an unknown date be indexed?
In a stated bucket for undated entries, visible from the period pages. Assigning them silently to a plausible period is how an index acquires structure that came from the indexer.
Can I let readers switch bucketing rules?
If you stored the rule per entry, yes, and it is the strongest version of this design. If you did not, the switch will quietly change which entries appear and nothing will record why.
