What science can tell us about a historical object

Before you begin

A laboratory can establish the origin of a raw material convincingly while leaving almost the object's entire biography open. There is no contradiction: each technique answers a bounded question. If an obsidian mirror has a composition consistent with a Mexican geological source, the finding can transform its study. It does not automatically supply the maker's name, purchase date, transport route or user's identity.

This lesson examines the British Museum mirror traditionally associated with John Dee, linking chemical analysis, collection records and historical attribution. A 2021 study by Stuart Campbell, Elizabeth Healey, Yaroslav Kuzmin and Michael D. Glascock compares its composition with geological samples. Understand the conclusion about Pachuca obsidian within that procedure, without extending it to questions requiring other sources. [1][2]

By the end you can describe what portable X-ray fluorescence measures, assess reference samples, distinguish precision from historical scope and separate material origin from documentary provenance in a catalogue entry. Your final evidence is a dossier with a bounded scientific conclusion and historical research plan. Assessment concerns explaining the evidence behind each claim, not memorising instrument names.

No previous chemistry knowledge is required. Concepts are introduced through examples and invented numerical datasets. These values do not reproduce the mirror's measurements and must not be cited as the article's results. Explanations describe professional research logic, not instructions for intervening on heritage objects. Where specialist conservation is needed, learn to recognise that need and specify the information sought.

The route begins with the object, moves through measurement and returns to history. Identify object and question, then examine calibration, repetition, references and comparison. Finally work through provenance, attribution, workshops and solutions. Sections can be consulted separately, but always distinguish observed data, analytical findings and historical interpretation. That distinction sustains the lesson.

1. Begin with the specific object

Before discussing science, identify the object studied. Popular names may refer to several objects, replicas or images. An inventory number links analysis to an institution's particular holding. The British Museum registers this mirror as 1966,1001.1. Its record brings together material, dimensions, association with Dee and collection information. Identification starts the enquiry; it does not prove every historical claim attached to the name. ^[1]

The museum description identifies obsidian and distinguishes associated elements such as the case. Collection objects can combine components of different periods and materials. Findings about analysed stone do not automatically apply to wood, leather or a label. Establish which part was measured: the inventory unit and analytical unit do not always coincide completely.

Describe before interpreting

An initial record should note form, attributed material, visible condition and institutional reference. Separate interpretations: function, cultural attribution, proposed date and association with a person. Each may be well supported but needs its own basis. Identifying it first as ‘a mirror used by Dee’ can turn attribution into an invisible premise. ‘A mirror associated with Dee’ preserves the documented relationship while leaving its basis open to examination.

The same caution applies to photographs. An image may show a dark reflective surface without determining its composition. Nor does it allow precise measurement without a suitable scale. Describe visible features and consult instrumental or catalogue information for other properties. Separate personal observation from catalogue information, noting each feature's source and your access.

Object, image and record

The physical object, its photograph and museum record are related but distinct sources. Photography selects lighting and angle; the record assembles information and classifications; analysis produces measurements under specified conditions. Discrepant published dimensions may require checking which part was measured and by what criteria before declaring an error. Identifying each representation and its function clarifies research.

You need neither handle the object nor take measurements. Read a study and institutional record critically, reconstructing relationships between object, procedure and conclusion. This skill supports evaluation of science reporting and museum materials without laboratory access. Indirect access does not prevent research, provided mediated information is not presented as personal observation.

First activity

Write a five-line record: identification, material, scientific question, historical association and access limit. An adequate answer identifies the mirror by inventory number, states that the study compares obsidian composition and distinguishes its association with Dee. Specify whether you consulted a catalogue record, article or image. Do not add a precise use date if these resources do not establish one.

This brief activity prevents escalating errors. Without identifying the analysed object, you cannot connect findings to its biography. Without knowing the measured component, you may extend findings to others. Confusing a catalogue name with ownership evidence makes chemistry appear to confirm a history already assumed. Careful identification underpins the dossier.

2. Two kinds of provenance: complementary, not interchangeable

Material provenance asks where a substance originated. Historical provenance concerns manufacture, circulation, ownership and changing meanings. These enquiries can connect but do not automatically answer the same questions. Stone may be extracted in one place, worked elsewhere and circulate widely. Its composition can retain a geological connection without recording subsequent transfers.

Imagine a fictional ring made with metal from one region and sold in another. Identifying the metal's origin does not identify the goldsmith; a sales invoice does not establish the mine. Securely connecting both kinds of evidence to the same ring improves its biography, but neither independently fills the other's gaps. The analogy separates questions without implying that all materials support the same provenance analyses.

A question matrix

The table assigns no disciplinary monopolies. Material traces can help investigate use, and documents can identify raw-material origins. Every claim needs a chain of support. Clear questions make historical and scientific collaboration productive. Neither ask laboratories to confirm whole narratives nor use documents to disregard inconvenient material findings.

Changing meanings

The object can retain its material and acquire different functions. A piece produced in one context can later become a collection object, an instrument attributed to a famous figure or an emblem of an exhibition. Composition does not determine a single meaning. To study those transformations, you need to identify who interprets the piece, when and through what documents. Material biography and the history of reception intersect, but are not equivalent.

For the mirror associated with Dee, chemical research provides a dimension of material origin that must remain alongside the documentary history of the attribution. The figure’s fame should not absorb all attention to the piece. Nor does the geological result allow the history of collection to be discarded without examination. The aim is to construct a broader description in which each dimension retains its evidence and unanswered questions. [1][2][3]

A writing test

Compare ‘chemistry proves Dee received the mirror from a Mexican merchant’ with ‘chemical comparison supports a provenance for the obsidian; transfer to Dee requires historical documentation’. The first introduces an agent and event absent from the measurement. The second preserves the finding and opens a separate enquiry. It avoids a narrative that appears scientific only because it follows a laboratory conclusion.

Organise separate columns for material, manufacture, circulation, ownership, use and reception. Some will hold firm information; others remain open. Do not fill every cell to create apparent completeness. A rigorous library can present a well-supported partial biography. Visual completeness matters less than honesty about what remains unknown.

3. What portable X-ray fluorescence does

X-ray fluorescence, abbreviated XRF, uses radiation's interaction with material to obtain elemental-composition information. In introductory terms, the instrument excites atoms and detects emitted radiation at characteristic energies. The observed pattern helps identify elements and, through evaluation and calibration, estimate concentrations. It does not read geographical names inside stone. It produces signals interpreted through models and references.

The initial letter of pXRF indicates that the equipment is portable. That feature can facilitate the study of objects that should not be transported or sampled invasively. But portability does not mean that there are no conditions of operation. Geometry, surface condition, material and procedure affect measurement. The result must be read alongside the method’s description. A photograph of the instrument beside the object does not replace that information.

From spectrum to concentration

A spectrum organises signals by energy. Features may correspond to elements; intensity and signal processing contribute to quantity estimates. Converting signals to concentrations does not directly count individual atoms in the object. Instrument response, measurement conditions and conversion models intervene. A concentration table is therefore a result developed from observations, not an immediate transcription of matter.

You need not master all the physics to evaluate the study's logic. Ask which elements are reported, in what units, with what calibration and repetition checks. Check whether the procedure can distinguish the compared materials. An instrument may measure several elements yet fail to resolve provenance if sources resemble one another too closely in those variables.

Surface and analysed volume

Treat ‘surface analysis’ carefully. Signals arise from a material zone whose effective depth depends partly on radiation and elements. They do not automatically describe the entire object uniformly. Surface layers, treatments or irregular geometry may affect particular readings. Ask whether conditions adequately represent the composition relevant to comparison.

In the mirror study, the authors describe repeated readings at two locations and note that they could not clean the surfaces of the museum objects. They also explain calibration and comparison with geological samples. These details allow the conditions of the result to be evaluated. Mentioning a limitation does not by itself invalidate the analysis; it identifies a factor that must be considered when interpreting the consistency of the measurements. ^[2]

An analogy with limits

Think of composition as a set of characteristics for comparing materials, just as several features distinguish similar objects. Avoid calling it an infallible individual fingerprint. Geological sources can vary internally and resemble other sources in some elements. Comparison relies on observation sets and separation criteria. The analogy fails if it implies each stone carries a unique, unambiguous label.

The historian's relevant conclusion follows evaluation of that comparison: which material origin is best supported, within what limits? The technique may work correctly yet be insufficient for other questions. It cannot identify an ownership date because that variable was not measured. Understanding the instrument reveals both its power and the boundary between findings and historical narrative.

4. Reference samples: the other half of the evidence

An isolated composition contains no geographical address. Proposed provenance requires comparison with materials of adequately known origin. Reference samples supply that relationship. Poor identification, insufficient samples or failure to represent a source's variation can weaken comparison. The reference set belongs to the argument, not to dispensable technical detail.

The mirror article compares materials from Mexican sources including Pachuca and Ucareo. Express findings in relation to the samples and variables used. It is not necessary to add generic doubt to every sentence, but retain the comparative framework. A strong match within a relevant set can support a firm attribution without becoming all-knowing identification of every object characteristic. ^[2]

Representing a source

Imagine a fictional geological source represented by one sample with a value of one hundred for an element. Its surrounding variation remains unknown: another area might produce different values. Several thoughtfully distributed samples characterise diversity better. More samples without attention to provenance and selection still do not guarantee representativeness. Sampling quality matters alongside quantity.

Now imagine sources A and B overlapping for one element. An intermediate reading cannot securely distinguish them using that variable alone. Another element may separate them better or preserve the overlap. Decisions depend on the combined pattern and uncertainty. Do not choose only the characteristic favouring your hypothesis and omit complicating evidence.

Geological and instrumental references

Distinguish two meanings of reference. Geological samples connect composition with source. Calibration or control reference materials assess measurement performance against established values. These functions may connect but are not identical. One supports origin attribution; the other contributes to measurement quality. A well-located comparison set does not remove the need for an evaluated instrument.

This distinction prevents assuming a study is sound because it uses many known stones without explaining how readings became comparable. Conversely, excellent calibration cannot compensate for an inadequate geological set. Findings require a complete chain. Identify how a weak link affects conclusions instead of judging the work by one signal of authority.

Questions when reading an article

Identify sample origins, numbers and comparison methods. Ask whether selected sources address the historical and material problem and whether the authors consider internal variation and alternatives. Acceptance does not require every geological source on earth. Assess the set's relevance and whether conclusions remain within its scope.

A sound synthesis might state: ‘The object's composition approaches one source's reference group and differs from analysed alternatives in the elements considered.’ Explain the implications for material provenance without adding a workshop or trade route absent other evidence. Geological references locate raw material; historical research must establish further links.

5. Repetition, precision and reference values

Repeated measurement helps assess consistency under defined conditions. Similar readings provide evidence of precision, but repetition alone does not guarantee closeness to a reference value. An instrument may produce very stable readings with a systematic offset. Similar numbers in a table do not fully validate a procedure. ^[4]

Consider an invented example. A control material has a reference value of one hundred units. One instrument reads one hundred and ten, one hundred and eleven, and one hundred and ten: close together but far from one hundred. Another reads ninety-nine, one hundred, and one hundred and one. Reference comparison reveals something repetition alone cannot. Do not call the first series ‘correct’ simply because it varies little.

What a mean contributes

A mean summarises readings without automatically removing their problems. Shared bias can persist; averaging materially different areas can conceal heterogeneity. Before summarising, ask what each reading represents and why combining them is legitimate. An average is useful when appropriate to measurement design, not a substitute for assessing conditions.

Three fictional readings of one hundred and one, one hundred and three, and one hundred and five have a mean of one hundred and three and an observed range of four. Describe that summary without turning the range into a complete uncertainty estimate. Repetition variability is one component; calibration, references and object conditions can contribute others.

Repeating one point and measuring several

Repeating measurements at the same location helps evaluate stability under similar conditions. Measuring different locations can show spatial variation or surface effects. Both operations are useful, but answer different questions. If all the points agree, consistency supports the interpretation within the design’s coverage. If they differ, you should not immediately average them to conceal the discrepancy. Investigate whether the difference has a material or instrumental cause.

Independence also matters. Many readings at one point are not many independent geological samples. More observations do not necessarily broaden material representation. Distinguish objects, points, repeats and samples. The previous lesson avoided counting proceedings as people; apply the same discipline to measurement units. Large counts need an explanation of what was repeated.

Proportionate language

Use ‘consistent readings’ when that is what has been checked. Use ‘comparison with reference values’ when that relationship has been evaluated. Do not turn either expression into ‘absolute certainty’. Metrology offers a vocabulary for describing measurement quality without demanding perfect results. Its usefulness lies in making conditions and limits explicit, so that other people can evaluate the conclusion. ^[4]

Historical readers should identify which control addresses which problem. Suspected surface variation prompts questions about measurement points; instrumental offset prompts calibration and control questions. An inadequate geological reference set is not repaired merely by repeating object measurements. Direct each question to the relevant research link.

6. Calibration: making signals comparable

Calibration establishes relationships between instrument indications and reference values under specified conditions. It is neither simply pressing a button nor declaring equipment modern. Distinguish calibration from instrument adjustment, although a procedure can connect them. You need not reproduce calibration to read a study, but should check how reported concentrations were obtained and evaluated. ^[4]

Two everyday scales may give different numbers for the same object. Comparing their weights requires knowing their relationship to references and conditions of use. Elemental analysis is more complex because several signals and material properties intervene. This analogy explains the need for comparability without describing every XRF mechanism.

An offset exercise

In a greatly simplified teaching model, an instrument systematically reports ten units above a reference across the relevant range. Ignoring that offset when comparing another series can create artificial differences. An adequately evaluated correction may improve comparison. Do not subtract ten from arbitrary real readings: this illustrates a problem, not a universal calibration rule.

Relationships may depend on element, concentration range and other conditions. Calibration evaluated for one set does not automatically transfer to every material. Read which references researchers used and how performance was checked. Many decimal places do not demonstrate adequate evaluation. Displayed digits and measurement quality are different properties.

Comparisons between instruments

If the object is measured with one instrument and the references come from another, you need reasons to consider them comparable. There may be calibration procedures and checks between series. The article should explain how that relationship is controlled. Do not assume that readings are incompatible because they come from different equipment, or that they are identical because they use the same units. Comparability is evaluated; it is not inferred from the table’s format.

High agreement in a check can be informative, provided you know the comparison it describes. It is not a probability that the object's whole history is correct. Instrumental fit concerns measurement; ownership and use belong to another level. Science reporting often makes this unjustified leap when summarising technical findings.

What to retain in your notes

Record method, units, calibration references and relevant controls in detail proportionate to your question. An introductory entry may summarise calibrated compositional comparison and cite the article; evaluating discrepant analyses requires more detail. Do not copy technical vocabulary without explaining its function. Precision means showing why it matters.

Distinguish whether instrumental response was evaluated, whether readings are consistent and whether geological samples represent compared sources. Each needs its own controls. Together they support material attribution, but none independently identifies a manufacture date. This separation values science without treating it as authority on every question.

7. Reading a multi-element comparison

Several sources may share one characteristic. Comparing multiple elements broadens the pattern, but longer lists do not automatically improve identification. Elements must be adequately measured and relevant; some discriminate poorly or are interrelated. Examine the whole set without selecting only features favouring a prior conclusion.

These fictional data use teaching concentration units and are not Pachuca, Ucareo or mirror values. Reference profiles are A: one hundred, twenty and ten for elements one, two and three; B: forty, eighty and thirty; C: sixty, thirty and eighty. Object O gives one hundred and three, twenty-one and eleven, approaching A in all three variables.

What the table permits and what is missing

Numerical proximity is visible, but dispersion, uncertainty and within-source variation are absent. Exact probabilities cannot be assigned. Establish whether differences are small or large relative to those conditions. One hundred and three may fit a group centred on one hundred if its variation permits; another context may demand caution. Distance needs interpretation as well as calculation.

We also do not know whether the references cover all relevant alternatives. Initially conclude ‘O is closest to A among the profiles shown’. Geological attribution needs suitable reference design and classification criteria. Four table rows teach comparison logic without reproducing the real study's complexity. Teaching simplifications must state omissions.

The role of units

Check matching units and expression bases before comparison. Parts per million are not percentages; a value expressed as an element cannot be mixed without conversion with differently expressed values. No conversions are required here, but recognise the question. Visually uniform tables may contain quantities that are not directly comparable; headings and methods must clarify this.

In a statistical comparison, scale can also influence the weight of each variable. An element with much larger numerical values can dominate a calculated distance unless it is treated appropriately. You should therefore not invent a criterion by adding up differences arbitrarily. Professional analysis justifies its procedures. As a reader, you can evaluate whether a conclusion rests on several consistent features without simulating a classification that you have not defined.

From resemblance to attribution

Observed resemblance is one stage. Attribution relates it to variation, uncertainty and alternatives. Clearly separated groups and consistent object agreement strengthen inference; overlap may require retaining several possibilities. Conclusions should be neither habitually weak nor enthusiastically strong: match the evidence's discriminatory capacity.

Reading a two-element diagram

A scatter plot positions each sample using two values, one on each axis. To read it, first identify the elements and units. Then distinguish the reference samples from the object through labels or symbols. If the object appears close to a group, observe what that proximity means in relation to the scale. A large visual separation can result from a very narrow axis; a small one can conceal significant differences if the axis spans an enormous range.

A two-element plot omits other variables used in a study. Plotting elements one and two in this exercise excludes the third. Use it for orientation, then return to the full table. Do not decide attribution from the most favourable image alone. Check other variables and measurement conditions against the interpretation.

Ask what each symbol represents. Source means conceal sample dispersion; multiple points may show it but could represent repeated measurements rather than independent samples. Legends must specify this. Investigate how group envelopes were constructed before treating them as provenance boundaries: they may be descriptive or model-based.

An accessible explanation of the figure could say: ‘Object O approaches A in the two elements shown; the table shows that it also does so in the third. The figure does not include the full uncertainty or demonstrate exclusive provenance by itself’. This description allows visual intuition to be used without confusing it with a complete classification. The chart’s aim is to make a relationship visible that you can then explain with words and data.

Visual reading: a bounded resemblance

Original teaching diagram. Object O approaches A in both plotted elements. The preceding table includes a third variable. These fictional values do not reproduce mirror data. Graphical proximity identifies neither workshop, date nor owner, and does not replace evaluation of uncertainty and references.

8. Uncertainty: what remains open

Measurement uncertainty does not mean every result is equally valid. It describes an evaluated dimension of the dispersion of values attributable to the measured quantity, within procedures and conditions. Historical enquiries also involve uncertainty about identification, provenance and context. Do not merge these under a vague word: each affects a different claim. ^[4]

You can have a consistent measurement and an uncertain historical attribution. You can also have a well-documented chain of ownership without knowing the material’s geological origin. Strength in one dimension does not automatically transfer to another. Organise the questions into levels to avoid that transfer: signal, concentration, reference group, geological source and biography. Each step adds a relationship that must be justified.

Distinguishing sources of uncertainty

Measurement can involve repetition, calibration and object conditions. Geological comparison involves reference variation and separation between alternatives. Historical identification asks whether documents describe the same object. Chronology may involve approximate dates or different events. Identifying uncertainty directs evidence searches: more chemical readings will not necessarily resolve inventory ambiguity.

In a fictional example, analysis clearly places a stone within group A, but an old inventory lists ‘a black stone’ without dimensions or distinctive features. Weak documentary linkage does not invalidate material findings; material findings do not make generic descriptions secure identifications. Preserve both assessments rather than declaring everything doubtful or everything proven.

Non-detection is not proof of absence

If an element is not detected in a way that provides useful information, you should not automatically conclude that its concentration is zero. Detection capability depends on the method and conditions. A report may use expressions or limits that need interpretation. The lesson does not aim to teach you to calculate them, but to recognise that ‘not detected’ and ‘absent’ are different assertions. The same discipline appeared when reading empty fields in a historical database.

Do not attribute every discrepancy to uncertainty to avoid revising a hypothesis. Readings substantially contradicting expected patterns need examination. They may expose measurement problems, inadequate references or mistaken attribution. Uncertainty does not accommodate arbitrary results: it supports proportionate evaluation of differences and compatibility using explainable criteria.

Numbers and language

Do not derive historical-certainty percentages from technical indicators. Calibration fit is not the probability Dee owned the mirror; small reading dispersion is not the probability of a particular workshop. Use evidence-based expressions such as ‘matches closely’, ‘supports’, ‘is consistent with’ or ‘does not resolve’, explaining their basis. Precise language can be more honest than unmodelled numbers.

Place concrete limits beside relevant conclusions: ‘The attribution concerns raw material; it does not establish the carving date.’ This is more useful than a generic warning about science's limits. Locate the inference boundary so readers can confidently use findings within scope and identify the research needed beyond it.

9. What the Pachuca finding contributes

The study by Campbell, Healey, Kuzmin and Glascock concludes that the composition of the mirror associated with Dee closely matches Pachuca obsidian in the comparison performed. This result provides a material basis for the discussion of provenance. It is not limited to a visual impression or the repetition of a catalogue attribution. It relates measurements of the object to geological samples through a described procedure. ^[2]

Assess that conclusion within its question. Adequate raw-material source identification may rule out or weaken incompatible hypotheses and guide production and circulation research. Guidance is not proof. Findings change historical possibilities without automatically supplying every link between geological source and museum.

A conclusion at three levels

First describe measurements and comparison; second attribute raw material to the best-supported source; third interpret historical consequences. Preserve these levels in summaries. ‘Science confirmed the mirror's entire history’ erases the actual question investigated. Brief responsible summaries retain composition, material attribution and biographical limits.

Pachuca identifies geological provenance relevant to the obsidian, not independently the exact carving location. Raw material can travel before working and finished objects afterwards. Reconstructing these stages requires archaeological, technical and documentary information. Composition connects to part of a material trajectory, not a complete itinerary inscribed in stone.

Why the finding still matters

Recognising the limit does not make the finding trivial. A raw material’s provenance can help situate an object within networks of availability and production, revise previous attributions or compare it with other pieces. It can open new questions about technology and circulation. Its value lies in resolving a question with appropriate evidence. It does not need to become a complete biography to justify its relevance.

Imagine competing fictional hypotheses of European local material or Mexican material. Relevant analysis supporting a Mexican source substantially changes discussion while leaving several acquisition routes and uses possible. Evidence reduces alternatives without necessarily producing one final narrative. Research advances through justified constraints and connections as well as complete accounts.

Compare without generalising to every object

Do not automatically give every dark mirror in a collection the same provenance. Similar objects may use different material sources. Formal resemblance guides comparison but does not replace object-specific analysis where material provenance requires it. Similar composition also does not guarantee joint manufacture: shared source, workshop and trajectory are different relationships.

When adding findings to a library, identify object, study and publication date. Separate new measurements from catalogue updates summarising them. Research dates are neither manufacture nor acquisition dates. This temporal discipline prevents mixing centuries of history with years of analysis. Science adds a documented layer to the history of knowledge about the object.

10. Raw material, workshop and manufacture date

Knowing a material's origin does not necessarily identify where it was worked. Workshop attribution needs production evidence, technical comparison or relevant documentation. One geological source can supply several places; one workshop can use several sources. Investigate the link rather than turning raw-material location into manufacture coordinates for narrative convenience.

Form and traces of working can provide complementary information. However, they must be studied with specialist criteria and appropriate comparisons. Formal resemblance can reflect a broad technical tradition, a later copy or similar functions. It is not enough to attribute a piece to a particular person. Historical analysis gains strength when several independent pieces of evidence converge on the same relationship, rather than when one observation is given different names and counted several times.

Material age is not object age

Rock may form long before extraction, carving, transport or use. Geological dates are not manufacture dates. Here compositional comparison does not independently give an exact carving date. Dating needs other evidence and identification of the event dated. ‘Ancient’ is too general to replace that distinction.

Imagine a fictional object recently carved from ancient material. Its composition can match a genuine geological source while manufacture is modern. This is not a claim about the mirror: it distinguishes material authenticity from manufacture chronology. Attribution research must address relevant dimensions, not just the easiest to measure.

Broad dates and caution marks

Catalogues may date objects using ranges or question marks. Preserve these when they affect confidence: a range spanning centuries must not become an exact year, nor cultural classification instrumental dating. Identify the basis: comparison, context, attribution tradition or analysis. Brief records need documentary interpretation.

The British Museum record presents a broad, cautious dating of the mirror, while the chemical analysis addresses material provenance. These pieces of information can coexist without one automatically resolving the other. If you write a chronology, put the proposed date of manufacture in a different row from the date of the study and the museum’s acquisition. Each event thereby retains its own evidence. ^[1]

Designing the next question

To investigate manufacture, seek comparisons of form, technique and find contexts alongside relevant documentation. For circulation, seek transport records, inventories and transfers. Such sources need not exist for every stage. Separate desirable from available evidence; good questions can acknowledge insufficient material for an answer.

Avoid reducing everything to composition: objects have histories of working and use. Equally, do not disregard composition because it cannot answer everything; it can decisively constrain hypotheses. Disciplines collaborate by assigning evidence functions and connecting findings through explicit arguments. Biography grows through justified connections, not automatic extrapolation.

11. Building an ownership chain

Historical provenance follows objects across contexts and owners through inventories, sales, correspondence and collection records. Identify every link: does it describe this object or a similar one? ‘Black mirror’ may be too generic. Dimensions, shape, marks, case and cross-references can help when assessed together.

Possible owners alone do not make a chain. Explain the document supporting each transfer and its confidence level. Chains may contain firm links, attributions and gaps. Identical arrows can falsely suggest proven continuity. Distinguish states through wording or symbols with accessible explanations. Design must not fill documentary gaps.

Identifying the same object

Two fictional inventories describe respectively a black disc with a pierced tab and dimensions, and a ‘seeing stone’. The first offers more distinguishing features but still needs context connecting it to today's object. The ambiguous second can matter when linked to other documents. Identification depends on discriminating features, not description length alone.

You must also consider material changes. A case may have been added later; a repair may alter the appearance; a measurement may include or exclude a projection. These possibilities should not be used as excuses to make any record fit. They are hypotheses that require support. If an important discrepancy remains unexplained, retain the uncertainty. Continuity of ownership should not be imposed because a narrative is attractive.

Contemporary documents and later memory

Collection labels may preserve important attributions, but their dates determine direct evidential scope. A label written long after a famous person's lifetime documents association when written. Earlier ownership requires investigating its basis, perhaps an earlier catalogue or oral tradition. Follow that chain rather than treating the label as contemporary testimony to the use described.

In the mirror’s history, the label associated with Horace Walpole constitutes one element of the attribution to Dee. The 2021 article places it within a broader documentary discussion. Its interest for this lesson is in showing that an object can carry a collection narrative with it. That narrative forms part of its history, but must be distinguished from chemical analysis and from direct proof of every earlier episode. [1][2]

Recording gaps without dramatising them

Documentary gaps neither disprove attribution nor justify accepting it unquestioningly. Assess support on either side and plausible alternatives. A historically supported association may retain unresolved aspects. Not every attribution fits complete proof or invention; vocabulary must express degrees and kinds of relationship.

Provenance records should include event, date, source, object identification and access level. Declare inventories known through scholarly quotation rather than claiming direct reading. Transparency lets researchers continue where checking remains. Ownership chains and chains of knowledge about them must be reconstructible separately.

12. John Dee: historical association and material evidence

John Dee’s name immediately directs the reading of the mirror. It can lead every feature to be interpreted as part of a personal practice. Research, however, needs to separate three questions: the object’s existence, its documentary association with Dee and the specific use attributed to it. Chemistry contributes mainly to the first, material dimension and to its provenance. It does not by itself identify the owner or confirm a ritual experience.

Silke Ackermann and Louise Devoy studied British Museum objects associated with Dee in 2012, examining provenance and documentation. Collections can combine objects with different attribution histories. One object's relationship does not prove those of others. Every object needs its own dossier, even when displayed together or under one thematic title. ^[3]

Possibility, opportunity and acquisition

A person may have been in places where similar objects circulated. That opportunity can make an acquisition hypothesis plausible, but does not demonstrate that they acquired this particular piece. To support the event, you need an additional link. The same rule applies to networks of contact: knowing a collector does not prove a transfer. The person’s biography can guide the search without taking the place of the missing document.

In a fictional situation, a scholar visited a city with stone merchants and was later associated with a mirror. A specific purchase cannot be deduced: gifts, other acquisitions or no ownership remain possible. Alternatives should not multiply indiscriminately, but show that opportunity does not determine events. Historical hypotheses must distinguish relevant possibilities.

Early descriptions can also be ambiguous

The 2021 article examines a 1624 account of a round, flat stone retrospectively connected with Dee through another person. It may support attribution discussion, but identification with the surviving mirror needs further reasoning. It is neither a chemical label nor an exact acquisition record: it has its own date, voice and scope. ^[2]

Value favourable evidence without exaggeration. Testimony closer to the period can strengthen plausibility, but temporal proximity removes neither descriptive ambiguity nor uncertainty about object identity. Ask what was seen, who asserted it and how it was transmitted. Connect those answers instead of treating ‘early source’ as a universal certainty seal.

Use and meaning

A reflective object can be used in different ways and receive different interpretations. Its physical capacity to reflect does not demonstrate a specific historical function. An attribution of use requires relevant context or documentation. Likewise, studying a belief does not mean confirming its supernatural effects. You can investigate what was expected of the object, how it was described and its role in an account without turning that expectation into a property demonstrated by material analysis.

When writing for students, preserve that difference through clear verbs. ‘Is associated’, ‘is described’, ‘is attributed’ and ‘the analysis supports’ are not interchangeable. Each indicates a relationship with the evidence. Rigorous prose can flow without eliminating these distinctions. The mirror’s history becomes richer when material, collection and reception remain visible, rather than being reduced to an instrument whose biography seems entirely known from its name.

13. Cultural biography without erasing original context

An object can become famous through its association with a European figure while retaining a material history that points to other contexts. Research must prevent later celebrity from absorbing the piece’s entire meaning. When studying the mirror, the obsidian’s provenance invites broader questions about production, circulation and transformations of interpretation. It does not, however, authorise the invention of a specific original function that the evidence does not demonstrate.

Attention to original context distinguishes material, cultural attribution and practice. A Mexican geological source does not automatically identify a producing community. A museum classification as Aztec needs historical and comparative grounding. General object–tradition relationships do not establish every specimen's use. These distinctions recognise non-European contexts without reducing them to homogeneous labels or undifferentiated mystical origins.

Circulation and transformation

When an object changes environment, its name, function and value may change. An inventory may treat it as a curiosity, an exhibition as an instrument or a study as material evidence. None of those readings exhausts its history. To study them, identify who is classifying it, at what moment and for what purpose. Cultural biography consists not in adding up possible meanings, but in documenting transformations in specific contexts.

You must also distinguish general context from a particular trajectory. The circulation of objects between the Americas and Europe after the conquest offers a historical framework. It does not demonstrate this piece’s exact route. You can use that framework to formulate questions and evaluate possibilities, but you need specific documentation to attribute a ship, merchant or date. Context makes a trajectory intelligible; it does not replace its links.

Documentary respect

Do not present regional heritage as an anonymous store of decorations for European history. Identify source terminology and consult specialist scholarship when extending interpretation. Without that basis, limit scope. Acknowledged gaps are better than generalisations about peoples or beliefs. Depth comes from defensible relationships, not unsupported cultural claims.

In a library entry, you can organise two routes: the earlier material and cultural history that can be investigated, and the later history of collection and reception. They do not have to occupy the same space if the available evidence is uneven, but both should be visible. Documentary inequality is itself information worth explaining. Silence in a European archive does not equate to an absence of history in earlier contexts.

A label-writing exercise

Compare ‘a mysterious object that reached a magician’ with a description identifying material, inventory number, attribution and provenance question. The first turns missing documentation into atmosphere; the second enables research. A concrete question about composition and collection history can arouse curiosity without exoticising uncertainty or making cultures scenery.

Aim for a plural, proportionate history. Science provides material relationships; documents explain some transfers and attributions; cultural comparisons open other questions. No layer should dominate through automatic prestige. Show connections and remaining separations, recognising the object's value beyond one owner, use or meaning.

14. Reading scientific articles without surrendering judgement

Historical readers need not become instrumental specialists to evaluate scientific argument structure. Identify question, object, method, references, findings and limits. This does not replace expert technical review but helps detect interpretative leaps and summarise faithfully. Understand what was done and which conclusions depend on it, avoiding automatic acceptance or rejection through unfamiliarity.

Start with the question and the identification of the objects. Then locate the method and the comparison samples. Only then read the conclusion as an answer to that design. If you start with a press headline, you may bring an overly broad expectation to the article. Reading in order helps separate what the authors measured from what a news report says they discovered. It also allows you to recognise when the article itself distinguishes hypotheses from results.

Six reading questions

Which particular object was analysed? Which property was measured? How was the measurement controlled? Which references were used for comparison? Which alternatives are better or less well supported? Which historical questions remain open? These questions organise the reading without requiring a summary of every paragraph. To answer them, use the corresponding sections and record the level of access. A partial view may suffice for some and be insufficient for others.

Peer review is not infallibility. Academic publication provides evaluation context, but arguments still need evidence and can change. Declared limitations do not automatically invalidate studies. Identify effects on a measurement, comparison set or subsequent interpretation. Useful criticism specifies location and consequence rather than invoking uncertainty universally.

Figures and tables

Read headings, units and captions before interpreting figures. Two-variable scatter plots need not represent entire classifications. Tables may show means while supplements provide further detail. Absence from an abstract is not absence from the study. Check accessible supporting material and declare unread parts. Evaluation must match what was consulted.

If a figure looks convincing, ask what observation would make it less clear: overlapping groups, additional references or greater uncertainty. This question does not require problems to be invented. It helps us understand what the argument depends on. You can also ask whether several figures provide different information or show the same data in related ways. Visual repetition does not automatically equate to independent confirmation.

Summarising within scope

Good summaries combine a main claim and boundary: ‘Analysis compares mirror composition with geological sources and supports material attribution; it does not independently establish acquisition date or route.’ Develop historical consequences with their own sources. Explain relevance without turning a bounded finding into total history.

Readers should know what you checked. State use of article and catalogue. If a claim comes from an earlier study cited in the article, preserve that mediation until consulting it. Concealing limits does not create rigour. Traceable reading is part of lesson quality: academic libraries teach results and responsible routes to them.

15. Worked workshop: from chemical table to conclusion

This workshop uses a fictional object, called disc O, and three invented geological sources. No number represents the mirror associated with Dee. The aim is to learn to read a comparison and state its scope. We have repeated readings, observed ranges in references and two ambiguous historical documents. You must keep those pieces of evidence separate before attempting to construct a biography. A chemical table and an inventory are not combined by simply adding certainties.

The teaching dossier

O's three readings for element one are one hundred and one, one hundred and three, and one hundred and five. For element two they are twenty, twenty-one and twenty-two. For element three they are ten, eleven and twelve. Their means are one hundred and three, twenty-one and eleven. In A's reference samples, we observe ranges of ninety-eight to one hundred and six, nineteen to twenty-three, and nine to thirteen, respectively. B and C present clearly different profiles in this simplified exercise.

Step one: describe before attributing

O's means fall within A's observed ranges in all three variables. They are outside B's and C's ranges as defined here. This describes the teaching dataset. We have not calculated probabilities or established that the ranges represent all the variation in each source. Nor have we yet included calibration uncertainty. The exercise permits an initial compatibility assessment, not a complete professional classification.

Step two: check the summary used

The mean for element one is obtained by adding one hundred and one, one hundred and three and one hundred and five, then dividing by three. The result is one hundred and three. That calculation is correct, but does not decide whether the three values should be grouped. We need to know whether they are comparable repetitions and whether conditions allow the object to be treated as sufficiently homogeneous for this question. The numerical datum and its interpretative legitimacy are connected but different questions.

The table’s reference ranges are the minima and maxima observed in the example. They are not confidence intervals or absolute limits of the sources. If we add samples, they might widen. You should not treat them as infallible natural boundaries. Their purpose is to show why variation matters when comparing a mean. A value close to another mean may have a different significance depending on the dispersion within each group.

Step three: state provisional material attribution

A proportionate conclusion would be: ‘In the available teaching data, O is compatible with A in all three variables and distinct from B and C. Material attribution to A requires accepting that the reference materials and procedure are suitable for the problem’. This sentence is not an evasion. It identifies the observed relationship and the conditions needed to turn it into an attribution. If those conditions were thoroughly checked in a real study, the conclusion could be expressed more strongly.

Step four: introduce historical documents

Now a fictional inventory from 1700 appears, mentioning ‘a dark disc’, together with an 1850 label attributing it to a famous person. Neither identifies the geological source on its own. Nor does the chemical table establish that the inventory's disc is O. We need features connecting description and object: dimensions, shape, marks or a collection history. Historical research begins where material identification stops answering the question.

The 1850 label documents an attribution at that moment. To use it as evidence of earlier ownership, we must know its basis. Perhaps it cites a lost inventory, perhaps it transmits a tradition and perhaps it contains an error. Do not choose an explanation without evidence. Chemistry can be compatible with the history and still not decide between those alternatives. Compatibility does not equate to confirmation of every detail.

Step five: write the final record

The record can use two separate paragraphs. The first presents material attribution to A under the exercise's conditions. The second explains that identification with the 1700 inventory and the ownership claimed by the label require further documentation. This organisation prevents readers from thinking the laboratory has validated the label. It also retains each item's value: measurement for material, inventory for a collection report and label for attribution history.

Extension: a new reference

Suppose a source D appears with ranges similar to A's in all three elements. Exclusive attribution to A weakens, even though O's measurements have not changed. The new problem lies in the reference dataset's ability to discriminate. Repeating the same variables a hundred times might improve precision while still failing to separate A from D. Other variables or evidence may be needed. The exercise shows why more measurements do not resolve every uncertainty.

The revised conclusion might say that O is compatible with A and D and that the current data do not distinguish them. This does not erase earlier work: it retains results and adjusts the scope in light of a relevant alternative. Well-documented research permits this change without rebuilding the entire chain from scratch. Quality includes the ability to revise an attribution when comparative evidence changes.

16. Audit workshop: correcting a scientific headline

Consider this fictional headline: ‘Science proves John Dee bought his mirror from a Pachuca workshop and used it in a particular ceremony.’ Below it appear an instrument photograph and a reference to the 2021 article. Separate claims and identify those within the analysis's question. Do more than call it exaggerated: locate and repair every leap.

Break down the claims

The headline combines material provenance, manufacture location, purchase, purchaser and use. Chemistry can support obsidian provenance but does not directly identify workshops, record purchase contracts or observe historical use. Additional claims need other sources. An article citation does not extend its evidence to everything following it.

Repair without emptying the finding

A better title is: ‘Obsidian analysis supports the material origin of the mirror associated with John Dee.’ Explain Pachuca and distinguish acquisition and use questions in the text. Preserve relevance without burying conclusions under generic doubt. Match the claim's verb and object to what was investigated.

If the article also includes documentary discussion, you can summarise it in a separate paragraph with its own bases. Do not attribute everything to chemistry. Interdisciplinary research brings together different evidence, and its conclusions can depend on combinations. Readers must know which part comes from measurement, which from inventories and which from interpretation. The word ‘science’ should not function as a container erasing those distinctions.

A second error: certainty through decimals

Another fictional report claims: ‘The laboratory confirms the biography with ninety-nine per cent precision.’ Ask what the figure represents: fit, comparison or unsupported journalism. Instrumental parameters are not ownership probabilities. Locate the original definition and identify changes in the headline. If untraceable, state that the consulted material does not justify it.

A third error: one object representing all

A label claims: ‘This analysis proves every mirror of this type came from the same workshop.’ It leaps from object to population and raw material to manufacture. Adequate object sampling and production evidence are required; formal resemblance does not supply them. Specify analysed objects and supported conclusions without exceeding design.

Partly guided activity

Review: ‘Because three readings agree, the historical attribution is certain.’ Identify what repetition shows, what further measurement control is needed and which evidence supports historical attribution. An adequate answer explains consistency under defined conditions, reference evaluation and ownership or identity documentation. Explain their different functions rather than merely listing them.

Independent work

Write a short news report with a headline, opening and a paragraph on limits. It must communicate the real case’s material result without inventing a route, workshop or use. Include a sentence explaining why the finding matters. Then check whether readers could confuse the date of the study with the date of the object. Good communication allows a conclusion to be understood while preserving its boundaries. It does not require every investigation to become a revelation resolving centuries of history.

17. An explained glossary of measurement and provenance

Obsidian

Volcanic glass whose composition supports geological-source comparison here. Photographs cannot identify every dark polished object as obsidian, nor does the material name determine function, date or cultural use. Establish material identification through relevant observation and analysis before using it to begin a biography.

X-ray fluorescence

A technique obtaining elemental information from signals emitted following material excitation. Signal-to-concentration conversion needs an evaluated procedure. It does not directly produce historical-provenance labels: findings require reference comparison and specific questions. Portability facilitates some settings without eliminating instrumental, material or conservation conditions.

Elemental composition

Description of elements and proportions within the method's measurement capacity, not a biography or full cultural identification. Interpretation needs units, conditions and comparison criteria. One matching element may discriminate poorly; several consistent elements may strengthen support. Relevance depends on alternatives and measurement quality.

Geological reference sample

Material of adequately known provenance used to characterise a comparison source. Value depends on identification, selection and representation of variation; one sample need not define an entire source. Distinguish geological references from calibration or control materials: one connects composition and origin, the other evaluates measurement.

Calibration

Establishing relationships between indications and reference values under defined conditions. Neither adjustment nor absolute correctness guarantee, it supports interpretation and comparability within procedures. Ask which references and evaluations were used. You need not reproduce operations to recognise concentration tables' dependence on this metrological relationship. ^[4]

Precision

Closeness of repeated measurement results under specified conditions. Distinguish it from closeness to a reference value and confidence in historical attribution. Stable readings can share systematic offsets, so assess repetition alongside other controls. State which quality aspect was demonstrated instead of making precision a synonym for truth. ^[4]

Measurement uncertainty

Characterises a dimension of dispersion of values attributable to the measured quantity using available information. It does not make all results equally valid. Distinguish identity and ownership uncertainty: each affects another argumentative link. Identify what remains open and evidence that could reduce it rather than combining all doubts into vague warnings. ^[4]

Discrimination between sources

The capacity of variables and references to distinguish relevant alternatives. Overlapping sources cannot automatically be separated through a match. More repeats can improve precision without resolving overlap; other variables or evidence may be needed. Attribution must explain why patterns favour one source, not merely identify resemblance.

Material provenance

Relationship between an object's substance and origin, investigated through suitable characterisation and comparison. It does not independently identify workshop, maker, owner or transport route. Raw material can circulate before and after working. Specify the analysed component and compared alternatives, preserving the distinction between material and biography.

Historical provenance

Documented reconstruction of manufacture, circulation, ownership and contexts within research scope, including firm sections and gaps. Each link must identify the same object across sources. Generic descriptions or later labels may inform without proving an entire chain. Record documents, dates, attributions and access levels so others can check links.

Attribution

A proposed relationship between an object and an author, owner, tradition, place or function. The relationship asserted and the evidence supporting it must be specified. An attribution of material is not an attribution of user. A piece’s conventional name can preserve a history of association without resolving every detail. Rigorous writing maintains that distinction and prevents the title from serving as proof of itself.

Object biography

Reconstruction of material transformations, movements and meanings over time, without implying every stage is known or one original function persists. It relates science, documents and reception while retaining distinctions. Identify documented episodes and hypotheses rather than filling gaps with pleasantly continuous narratives.

18. Retrieval questions and explained answers

What the instrument measures

Does portable X-ray fluorescence directly identify a piece’s former owner? No. It obtains information about elemental composition through a measurement procedure. That information can help investigate material provenance when compared with references. Ownership needs other evidence. A complete answer explains the intermediate step of comparison and avoids presenting the instrument as an automatic reader of geography, chronology or biography.

Why references are needed

Are several elemental concentrations enough to name a geological source? Not alone. Relevant comparison sets and evaluation of variation, uncertainty and alternatives are required. Isolated values contain no place names. Distinguish object measurement from source characterisation: both support attribution, and weakness in either can limit scope.

Very similar readings

Do three closely agreeing readings prove absence of instrumental bias? No. Consistency informs precision under defined conditions, but readings can share offsets. Controls and references assess other aspects; repetition also does not prove historical ownership. Separate measurement quality from interpretative scope while valuing repetition's contribution.

Material and workshop

Does obsidian matching a particular source locate its workshop there? Not automatically: raw material can move before working. Manufacture attribution needs additional evidence, such as production context or relevant technical comparison. This does not mean the workshop was necessarily distant, only that composition alone cannot establish the relationship. Compatible possibilities still need arguments.

Date and geology

Does geological origin independently give the carving year? No. Formation, extraction, manufacture, acquisition and analysis are separate events. This compositional comparison does not date them exactly. A strong answer proposes useful dating evidence and specifies the event to date. ‘Antiquity’ must not obscure temporal differences.

A later label

What does a collection label attributing an object to a famous person document directly? That attribution's existence in the label's time and context, if identification is checked. Earlier ownership requires examining its basis and object identity. Valuable evidence is not automatically contemporary observation of use. Chemistry changes neither documentary date nor voice.

A new source

When another source fits the same values, do previous measurements cease to exist? No. Exclusive-attribution capacity changes, not necessarily object data. Revise conclusions and seek distinguishing variables or evidence. Findings depend on comparison sets; revising hypotheses can preserve much knowledge already gained.

A spectacular news report

A headline claims analysis proves a particular sea route. Check historical documentation connecting object and route, and whether the article distinguishes it from material analysis. Composition records no ship names. Identify the leap and propose wording preserving genuine findings rather than merely alleging exaggeration. Criticism should improve communication of what is known.

19. Final project: material dossier and documented biography

Prepare a library entry on a historical object studied scientifically, using this mirror or another sufficiently documented case. Distinguish identification, analytical question, procedure, findings and historical consequences. Submit an accessible synthesis with supporting apparatus enabling claim checks. Assessment concerns responsible reading and use of analytical evidence, not performing physical analysis.

Bound the question

Choose questions matching available materials: how composition changes provenance attribution, or how material findings relate to collection chains. Do not promise complete biography from a brief catalogue and scientific abstract. Depth means developing a documented relationship well. Limited scope can produce valuable library work when evidence is clear.

Evidence inventory

Create one table for object, identifier, analysed component, measured property, method, references and findings; another for historical events, dates, documents and confidence in object identification. Initially separate them to prevent measurement absorbing history or narrative silently conditioning findings. Connect rows later with explicit arguments, not merely because columns share words such as origin.

Record the level of access to each source. A photograph, a catalogue, a partial view and a complete article permit different questions. If an assertion depends on a document cited by another work, preserve that mediation. Transparency does not diminish the project’s value; it allows its scope to be evaluated and the research to continue. Never write that you have observed a physical mark if you have only read another author’s description of it.

Evaluate the analytical chain

Explain the technique accessibly and identify at least two relevant controls or conditions, showing the problem each addresses. Repetition and reference comparison have different functions. Listing calibration or uncertainty is insufficient: relate them to conclusions. Readers should understand why findings do not emerge automatically from instrument names or precise-looking numbers.

Then formulate the material attribution in a sentence with a defined scope. Add which historical assertion would be excessive if based on that attribution alone. This pair of sentences demonstrates that you can communicate a result and its boundary. Do not write a list of limitations disconnected from the finding. Criticism should retain its relevance and show what additional research would allow progress. A good record offers positive knowledge alongside unresolved questions.

Writing and revision

Organise synthesis into object and question, material findings and documentary history. Chronologies should separate proposed manufacture, transfers, museum acquisition and analysis. Diagrams should distinguish firm and hypothetical links in words as well as visual form. Accessibility requires more than colour or line differences. The product should be readable and citable outside class.

Checking by another reader

Give the synthesis to someone unfamiliar with the case and ask what analysis established and what remains open. If they attribute an unintended purchase date or user identity to the laboratory, revise order and verbs. Authorial intention is insufficient where presentation enables misleading inference. Quality control includes readers' understanding.

Keep a final note on the next useful check: locating an inventory, comparing an earlier catalogue or studying geological references further. Tie action to a specific dossier limit. The entry can then grow with new evidence. Research produces a revisable structure for knowledge rather than a closed narrative.

20. Integrating science and history without evidential leaps

Science contributes most strongly when its question is defined. The Dee-associated mirror's composition supports relevant material attribution; collection history and personal association additionally require documents and criticism. Separating these does not artificially divide the object: it connects dimensions without demanding every answer from one test.

A chain you can explain

Begin with an identified object and property to measure, followed by procedure, controls and references. Compare and interpret findings, then connect other evidence to build history. Explain what each step retains and adds. Whenever a conclusion suddenly introduces a person, manufacture location or journey, identify evidence for that new link.

The chain can also show disagreements. A material result can complicate a catalogue attribution; a document can question the identification of a piece. Do not resolve the tension by automatically choosing the discipline you regard as more prestigious. Examine objects, methods and sources. Perhaps different components were compared or an ambiguous description was used. Rigorous collaboration begins by locating the discrepancy and formulating what check could resolve it.

Avoid two simplifications

One claims science reveals an object's whole truth at once; the other claims interpretative limits prevent firm knowledge. Both lose evidential structure. Strong material attribution can coexist with unknown routes; documented museum acquisition with uncertain earlier ownership. Distribute confidence appropriately rather than spreading certainty or doubt across the entire dossier.

Study libraries should teach this distribution: what readers can use as a basis, what depends on hypotheses and which documents permit progress. Tables, diagrams and glossaries help preserve relationships but cannot replace arguments. Editorial quality makes complex reasoning consultable without eliminating distinctions essential to correctness.

Revisit earlier lessons

The connection with modern titles, foundation dates and historical databases is profound. A title can anticipate an interpretation; a date can summarise several beginnings; a row can bring together a complex history; a composition can support a material origin. In every case, a useful representation risks being confused with the whole object. TMF’s method consists in identifying the mediation and reconstructing the step from evidence to assertion.

To close your study, write two sentences. The first should express the strongest result you can defend. The second should identify the most important historical question that the result does not resolve. If both are precise, you have learnt to use evidence without diminishing its value or exaggerating its scope. That competence allows you to continue researching with a firm foundation and a clear agenda, keeping object, measurement, document and interpretation together.

21. Sources and further reading

[1] British Museum. Record for the obsidian mirror associated with John Dee, number 1966,1001.1. Identification, material, proposed dating and collection information. Distinguish catalogue information from instrumental findings and hypotheses about earlier episodes. Accessed: October 2026.

https://www.britishmuseum.org/collection/object/H_1966-1001-1

[2] Campbell, Stuart; Elizabeth Healey; Yaroslav Kuzmin; Michael D. Glascock. ‘The mirror, the magus and more: reflections on John Dee’s obsidian mirror’. Antiquity, 95, 384, 2021, pp. 1547–1564. DOI: 10.15184/aqy.2021.132. Open-access article combining material analysis and historical discussion. This lesson's original, fictional numerical tables and examples do not reproduce its measurements.

https://doi.org/10.15184/aqy.2021.132

[3] Ackermann, Silke; Louise Devoy. ‘“The Lord of the smoking mirror”: Objects associated with John Dee in the British Museum’. Studies in History and Philosophy of Science Part A, 43, 3, 2012, pp. 539–549. Research into provenance of Dee-associated objects. This lesson uses bibliographical identification, the available abstract and links to later discussion; thorough investigation requires the full article.

https://www.sciencedirect.com/science/article/pii/S0039368111001117

[4] Joint Committee for Guides in Metrology. International Vocabulary of Metrology, JCGM 200:2012, and annotated VIM. Terminological reference for precision, repeatability, calibration and uncertainty. Lesson explanations and exercises are an educational introduction, not an instrumental protocol or substitute for professional evaluation.

https://www.bipm.org/en/committees/jc/jcgm/publications

https://jcgm.bipm.org/vim/en/2.15.html

https://jcgm.bipm.org/vim/en/2.39.html

https://jcgm.bipm.org/vim/en/2.26.html

Reading route

Start by identifying the piece in the catalogue. Then read the question, method, comparison and conclusion in the article, preserving their differences. Use metrological vocabulary when you need to make a term precise. To investigate ownership and use, return to the documentary discussion and its references. The exercises with A, B, C, D and O are fictional learning materials and must not be used to attribute real objects or infer chemical values for the mirror.

Sources for this reading

1. British Museum. Record for the obsidian mirror associated with John Dee, number 1966,1001.1. Identification, material, proposed dating and collection information. Distinguish catalogue information from instrumental findings and hypotheses about earlier episodes. Accessed: October 2026. Consult the source
2. Campbell, Stuart; Elizabeth Healey; Yaroslav Kuzmin; Michael D. Glascock. ‘The mirror, the magus and more: reflections on John Dee’s obsidian mirror’. Antiquity, 95, 384, 2021, pp. 1547–1564. DOI: 10.15184/aqy.2021.132. Open-access article combining material analysis and historical discussion. This lesson's original, fictional numerical tables and examples do not reproduce its measurements. Consult the source
3. Ackermann, Silke; Louise Devoy. ‘“The Lord of the smoking mirror”: Objects associated with John Dee in the British Museum’. Studies in History and Philosophy of Science Part A, 43, 3, 2012, pp. 539–549. Research into provenance of Dee-associated objects. This lesson uses bibliographical identification, the available abstract and links to later discussion; thorough investigation requires the full article. Consult the source
4. Joint Committee for Guides in Metrology. International Vocabulary of Metrology, JCGM 200:2012, and annotated VIM. Terminological reference for precision, repeatability, calibration and uncertainty. Lesson explanations and exercises are an educational introduction, not an instrumental protocol or substitute for professional evaluation. Consult the source