The gold standard of history of museums is explaining how collections became institutions, how their purposes changed, and whose knowledge and experiences they represented. Dates and famous buildings matter, but a useful history also follows objects, collectors, workers, visitors and the decisions that made some things visible while leaving others outside the display.
This guide teaches you to investigate those changes through verified institutional examples, source comparisons and original worked exercises. Its invented Harbour Museum cases are teaching scenarios, not accounts of an actual museum. The aim is to help you read a museum as both a place that explains history and an institution with a history of its own.
Begin with the Gold Standard master, connect the method with Education: Learning That Transfers into Life, or explore the How X Works hub. For present-day collection responsibilities, use Museum Studies. Here our main question is how those responsibilities and institutions developed over time.
Choose your museum history route
Chapters 1–3: Find origins and public purposes
Chapters 4–6: Follow collections and descriptions
Chapters 7–9: Read displays and institutional work
Chapters 10–12: Test public claims and change
Chapter index
Find origins and public purposes · Chapters 1–3
1. Start with a historical question, not a list of museums
Follow collections and descriptions · Chapters 4–6
4. Use Singapore to see an institution change its purpose
Read displays and institutional work · Chapters 7–9
7. Reconstruct an exhibition as an argument
8. Measure representation without confusing objects and people
Test public claims and change · Chapters 10–12
10. Test claims about visitors and learning
11. Explain continuity through relocation, loss and reorganisation
12. Recognise changing authority and community participation
Transfer and investigate · Chapters 13–14
13. Transfer the method to a digital exhibition
14. Build a practical research route and answer common questions
CHAPTER 1 OF 14 · Find origins and public purposes
1. Start with a historical question, not a list of museums
A museum brings several activities into a lasting arrangement: collecting, keeping, studying, interpreting and displaying material or other forms of heritage. Different institutions have combined those activities differently. Some collections supported scholarly work before they supported regular public visits. Others grew around public display, civic identity, teaching or a particular community’s wish to preserve its experience.
The first historical task is to identify which arrangement you are studying. A private collection, temporary exhibition, university museum and national museum may contain similar objects but operate under different rules. Calling all of them museums without examining those rules can hide the very change you want to explain. Ask who controlled the collection, what purpose they stated, how it was funded and who could use it.
The ICOM museum definition adopted in 2022 includes research, care, interpretation, public access and community participation. It is a useful professional reference for our own time. It is also a dated historical document. Do not assume that an eighteenth-century collector understood the museum’s duties in exactly the same way, or that adopting a definition automatically changes every institution’s practice.
Turn a topic into an investigation
“Old museums” is a topic. “Why did a local collection begin displaying industrial tools alongside paintings?” is a research question. It directs you towards acquisition registers, committee minutes, exhibition catalogues and visitors’ responses. It also makes competing explanations possible: a new curator, a changed collecting policy, a large donation or a desire to reach a different audience.
For the fictional Harbour Museum, suppose a catalogue from 1900 lists mostly portraits while one from 1930 includes machinery and workers’ possessions. You have evidence that the catalogues differ. You do not yet know whether the collection changed, whether the later catalogue describes previously unlisted holdings, or whether the publication served a different exhibition. The question becomes sharper as you separate those possibilities.
A sound account explains a change with a sequence and a mechanism. “The museum modernised” is too vague. “A revised collecting policy admitted workplace material, and the next two accession registers record donations from local workshops” identifies an institutional decision and evidence of its consequences. You can then investigate what the new exhibits meant to visitors.
CHAPTER 2 OF 14 · Find origins and public purposes
2. Understand collecting before the modern museum
Objects can be valued for devotion, memory, scholarship, status, beauty, rarity, teaching or practical reference. A history of museums should investigate these purposes without pretending that every earlier collection was simply waiting to become a modern gallery. The same object may move between several kinds of collection and acquire different meanings in each setting.
The Ashmolean’s institutional history traces its collection through the Tradescants and Elias Ashmole, whose gift to Oxford preceded the museum’s opening in 1683. It describes a collection containing natural specimens and human-made objects, and a building that also accommodated scientific teaching and a chemistry laboratory. This example connects collecting with investigation and instruction, rather than treating display as the only original function.
The lesson is a method, not a universal origin story. Trace the specific relationships through which objects were assembled. Who travelled, purchased, exchanged, received gifts, prepared specimens or supplied information? A collector’s name can stand at the top of a catalogue while concealing many other contributors. Recover those people where records permit, and acknowledge uncertainty where they do not.
Follow the change in meaning
Imagine a fictional navigational instrument that first served on a working vessel, then entered a retired captain’s collection, and later appeared in Harbour Museum’s maritime gallery. Its physical form might remain broadly similar while its role changes from equipment to personal keepsake to historical evidence. Each transition creates different questions and different kinds of documentation.
A purchase receipt might establish one transfer. A diary might explain why the captain kept it. A museum register might show how the institution classified it. None of those sources alone establishes its entire life. The instrument’s scratches may suggest use, but they cannot identify a particular voyage without supporting evidence.
Avoid a simple ladder from irrational curiosity to rational science. Earlier collecting could involve careful observation; later museums could preserve questionable assumptions. Compare the actual practices: naming, grouping, measuring, documenting and making claims. Historical change becomes clearer when you ask how knowledge was produced rather than assigning an entire period one flattering or dismissive label.
CHAPTER 3 OF 14 · Find origins and public purposes
3. Distinguish founding, opening and becoming public
Institutions often have several meaningful beginning dates. A collection is offered, a legal body is established, premises are obtained, staff are appointed, and visitors are admitted. The building that people recognise today may open much later. A single founding year can be convenient, but it must not flatten these events into one moment.
The British Museum’s account of its beginnings distinguishes its foundation in 1753 from opening its doors in 1759. That six-year interval is enough to show why “founded” and “opened” should not be treated as interchangeable. The relevant source for one claim may not establish the other.
The Louvre’s institutional overview likewise distinguishes the palace’s long architectural history from the museum inaugurated in 1793, following a 1791 decision during the French Revolution. An older building does not make the museum institution equally old. The conversion of a palace into a museum also directs attention to changing authority over collections and their public purpose.
Work an original timeline problem
Harbour Museum’s invented records include a donation agreement dated 1880, a governing committee established in 1882, an exhibition notice from 1884 and the opening of a permanent building in 1891. A heritage panel says the museum “opened in 1880”. What should be repaired? Keep the donation date, but describe it as a donation. State separately what the records establish about the exhibition and permanent building.
The agreement does not prove that objects arrived that day. The exhibition notice announces intended access but may need corroboration from reports of the event. A building photograph proves a physical structure existed when photographed; it does not prove every gallery was operating. Assign each source a specific evidential job.
Then investigate what “public” meant. Were visitors required to apply, attend at set times, join a guided group, pay or provide a recommendation? A policy of public admission can coexist with practical restrictions. This does not make the policy meaningless. It means that a historical explanation should identify the opportunities it created and the barriers it retained.
CHAPTER 4 OF 14 · Follow collections and descriptions
4. Use Singapore to see an institution change its purpose
Singapore offers a particularly useful example because museum, library and natural history collections share parts of their institutional history without remaining one organisation forever. Following those connections helps explain how collections can retain continuity while buildings, names, administrative homes and public purposes change.
The National Museum’s official milestones identify an 1823 proposal, an 1849 gift of two ancient gold coins from Temenggong Ibrahim of Johor, the 1874 government takeover and renaming, and the opening of the Stamford Road building on 12 October 1887. These are different events, not rival answers that must be reduced to a single birthday.
The same timeline records a later change of mandate in 1972 towards history, ethnography and art, alongside the movement of natural history collections. The Lee Kong Chian Natural History Museum’s history specifically traces the zoological collections’ move to the university in that year. Read those accounts together: a change in one museum’s focus can be connected to continuing research responsibilities elsewhere.
Trace institutions and collections separately
Draw two timelines. On one, record the museum’s name, authority and buildings. On the other, record a particular collection’s custody, location, research use and public display. Where they intersect, explain the relationship. Where they diverge, resist the temptation to call the collection lost merely because it no longer appears in the same gallery.
This approach repairs a common misconception: that a museum’s current theme describes everything it has ever held. A visitor who sees mainly social history today may not realise how important natural history was to its earlier work. Conversely, a familiar specimen may have a much longer institutional history than its present building.
For a focused learning route, read how natural history collections changed science in Singapore. The related History of Libraries root helps you investigate the library side of the shared story. Their connection is historical; neither should be treated as a substitute for the other’s complete institutional chronology.
CHAPTER 5 OF 14 · Follow collections and descriptions
5. Investigate how objects entered collections
An object does not arrive in a museum simply because it is historically important. Someone identifies, obtains, transports, offers, selects and records it. Those actions can involve gifts, purchases, excavation, exchange, official transfer or coercion. A history of collecting must examine the circumstances of each acquisition instead of treating every entry in a register as the same kind of event.
Begin with a distinction between evidence of receipt and evidence of the earlier history. An accession register may reliably show that the museum received an object from a named dealer in a particular year. It may say little about how that dealer obtained it. A precise final transaction does not automatically fill earlier gaps.
The Ashmolean’s own historical account acknowledges the role of colonial power in the formation of parts of its collection. That observation creates a research obligation, not a licence to invent an identical story for every object. Investigate the actors, relationships and records in the case before you. Power can shape access to objects and the language used to describe their acquisition.
Build an evidence chain
In an invented Harbour case, a 1910 register says a carved box was purchased from a merchant. A 1908 photograph shows a similar box in another collection, but no identifying mark is visible. A later catalogue treats the photograph as proof of the museum object’s previous owner. That conclusion is too strong. Similarity is a lead; identity needs additional support.
A better entry separates the facts: purchase recorded in 1910; earlier ownership unresolved; possible photographic comparison requiring investigation. If measurements, distinctive damage or correspondence later establish a match, the record can be revised with its reasons. Research advances through explicit corrections, not through quietly changing uncertain words into certain ones.
Questions about return, ownership and appropriate stewardship require case-specific evidence and engagement with relevant communities and authorities. A historical guide cannot settle every claim through a general slogan. It can show why incomplete records, unequal power and changing institutional responsibilities matter, and why possession, documentation and moral legitimacy should not be collapsed into one claim.
CHAPTER 6 OF 14 · Follow collections and descriptions
6. Read catalogues as changing systems of knowledge
A catalogue is both a source about objects and a source about the institution that described them. Its categories reveal what curators considered worth distinguishing. Its omissions may reflect available knowledge, collecting priorities, publication space or assumptions about the reader. The descriptions can change even when the physical objects remain in the same storeroom.
Compare editions carefully. Preserve the exact date and wording of each description, then identify the change. Was an object reassigned to a different maker, given a new date, described in another language or moved to a different category? Those revisions may indicate new evidence, new expertise or a change in the questions researchers ask.
The Historical Museum Catalogues Quick Notes guide offers a practical introduction to separating an object’s creation date, acquisition date and catalogue interpretation. For museum history, add another layer: why did this institution choose that description at that time?
Solve an identifier problem
Harbour’s fictional 1905 catalogue lists a ceramic vessel as item 42. A 1950 catalogue calls a similar vessel item 116. A modern database lists accession H.1903.7 and includes both earlier catalogue numbers. There may be one object with three identifiers, not three objects. Check the measurements, photographs, distinctive features and explicit cross-references before adding them together.
Now imagine the modern record contains two components: vessel and lid. Counting database rows might produce two, while the old catalogue counted the pair as one entry. Neither count is necessarily wrong, but they answer different questions. A historian who reports that the collection doubled from one to two has mistaken a descriptive change for an acquisition.
Retain historical terminology when necessary to explain the source, but contextualise it. Categories can preserve discredited assumptions or language that misrepresents people. Quoting a catalogue to analyse its worldview is different from adopting that worldview. Explain what the term did within the institution’s classification and how later descriptions revised or challenged it.
CHAPTER 7 OF 14 · Read displays and institutional work
7. Reconstruct an exhibition as an argument
An exhibition selects objects, places them in an order, supplies text and asks visitors to notice relationships. Its arrangement therefore makes an argument, even when the labels appear purely descriptive. A row of objects ordered by date suggests development over time. Objects grouped by material suggest a different comparison. A room organised around one ruler directs attention differently from a room organised around households or labour.
To reconstruct an older exhibition, use floor plans, installation photographs, labels, catalogues, guidebooks and correspondence. Each gives a partial view. A catalogue may include works that were not all displayed simultaneously. A photograph may omit a wall or record an unfinished installation. A floor plan establishes intended spatial relationships, not necessarily how every visitor moved.
Ask what the visitor encountered first, what received the most space, what was easy to read and what required specialist knowledge. An object placed centrally with a large label can acquire prominence through display. That prominence is evidence of curatorial emphasis, not automatic proof that the object was historically the most important item in the collection.
Compare two fictional displays
Harbour’s 1920 gallery puts a factory owner’s portrait beside a large machine. Its 1980 successor adds workers’ tools, pay records and oral testimony. The change could broaden the explanation from ownership and invention to labour and everyday experience. To support that argument, compare the actual labels and selection records, rather than inferring the whole message from two photographs.
Also investigate continuity. The later exhibition may still retain the owner’s portrait and machine, but give them a different relationship to other evidence. Historical interpretation often changes through rearrangement and contextualisation, not simply through replacing all earlier objects.
For a visitor exercise, write two accurate labels for the same invented object: one about manufacturing technique and one about domestic use. Compare which evidence each needs. The exercise shows how different questions produce different interpretations while still requiring factual discipline. A compelling story is not permission to supply a maker, owner or experience that the evidence does not establish.
CHAPTER 8 OF 14 · Read displays and institutional work
8. Measure representation without confusing objects and people
Collections are selected samples of surviving material, not miniature copies of the entire past. Some materials survive better than others. Some people have more opportunity to donate, sell or preserve possessions. Some activities generate durable objects and written records; others leave fewer traces. Collecting policies introduce another layer of selection.
The existing guide on why museums collect everyday objects explains how ordinary possessions can support historical inquiry. Their presence can widen the questions a collection answers. However, collecting an everyday object does not by itself establish that every group’s experience is adequately represented.
Work a selection calculation
Suppose Harbour has 1,000 catalogued objects: 700 from wealthy households, 200 from workplaces and 100 from other contexts. The first category represents 70 percent of objects in this invented collection. It does not follow that wealthy households were 70 percent of the town’s population, or that they produced 70 percent of its historical activity. The denominator is objects collected, not people who lived there.
The museum then acquires 200 workplace objects. Its holdings rise to 1,200, and the wealthy-household share falls to 700 divided by 1,200, approximately 58.3 percent. Nothing has been removed from that category. Its proportional share falls because another category grew. Explain both the counts and the denominator when evaluating a collecting change.
Even the revised proportions do not settle representativeness. Two hundred objects might come from one workshop, while a single recorded interview could illuminate experiences shared across several industries. Quantity is one dimension of coverage; diversity, provenance, context and relevance to the question are others.
A responsible conclusion names the gap precisely. “The catalogue contains no identified objects from this neighbourhood during the period examined” is stronger than “the neighbourhood had no history”. It is also more useful: it tells a researcher what the present records do not show and where further investigation might begin.
CHAPTER 9 OF 14 · Read displays and institutional work
9. Follow the labour and money behind public galleries
Founders and buildings often dominate institutional histories because they are easy to name and commemorate. Daily museum work can be harder to recover. Preparing specimens, registering acquisitions, translating labels, cleaning spaces, maintaining stores, answering enquiries and arranging transport all affect what an institution can preserve and explain.
Read accounts alongside staff records and work instructions. A new gallery may be funded by a large one-off gift, while the people needed to maintain it depend on a recurring budget. A public opening can therefore be a genuine achievement and still create new obligations that the institution has not fully funded.
Calculate a fictional operating gap
Harbour receives 60,000 accounting units for a new exhibition. It spends 40,000 on cases, 12,000 on installation and 8,000 on interpretation. The opening budget balances. Annual staffing, utilities and routine upkeep subsequently cost 18,000 units, but only 12,000 additional annual income is available. The recurring gap is 6,000 units per year. The original gift does not solve that continuing shortfall once spent.
This is a teaching calculation, not a historical cost estimate. In a real account, check whether expenses were paid centrally, donated in kind or assigned to another department. An apparently cheap gallery might depend on unpaid work or shared facilities. Do not infer the absence of labour from the absence of a separate budget line.
Staffing numbers also need interpretation. Five named employees may not mean five full-time posts. A curator might divide time between collections, teaching and administration. To explain a backlog or service change, reconstruct duties and available time instead of treating headcount as a complete measure of capacity.
Personal testimony can reveal practices absent from formal reports, but dates and responsibilities still need checking. A former employee may remember an informal routine more accurately than an official manual describes it, while compressing several periods into one memory. Treat disagreement as something to investigate. A good history can recognise skill and commitment while documenting institutional constraints and mistakes.
CHAPTER 10 OF 14 · Test public claims and change
10. Test claims about visitors and learning
Museums often explain their public value through attendance, educational programmes and visitors’ experiences. These measures matter, but they are not interchangeable. An admission count records entry events. A school booking records an organised visit. A questionnaire records what particular respondents reported. None automatically establishes lasting learning across the entire audience.
Before comparing years, check counting rules, opening periods and temporary closures. Did staff count visits to every gallery separately? Were event attendees included? Did an electronic counter replace a manual tally? Changes in measurement can produce a discontinuity that looks like a change in public interest.
Work a visitor comparison
In the fictional Harbour records, attendance rises from 24,000 to 30,000 visits. That is an increase of 6,000, or 25 percent. But open days rise from 200 to 250. Average visits per open day are 120 in both years. Total recorded use increased, while daily average use remained unchanged. Both statements can be true and should be reported together.
Suppose a survey then finds that 48 of 60 respondents can explain the main exhibition idea. The observed proportion is 80 percent among respondents. It is not automatically 80 percent of all visitors. People who stopped to answer may differ from those who left, and the question may test immediate recall rather than later understanding.
For historical research, visitors’ letters, diaries, press accounts and school reports can illuminate experience. Read them with attention to selection. A published complaint may reveal an obstacle, but it does not establish how common that obstacle was. A glowing opening review may reflect a special event rather than ordinary conditions.
Link claims to the evidence they actually require. If the question concerns access, investigate cost, travel, opening times and eligibility. If it concerns interpretation, examine what visitors encountered. If it concerns learning, look for evidence of what people understood or could do afterwards. This prevents impressive attendance totals from carrying conclusions they cannot support.
CHAPTER 11 OF 14 · Test public claims and change
11. Explain continuity through relocation, loss and reorganisation
Institutional change is not always visible as growth. Collections can move, split, merge, disappear from display or enter long periods of storage. A gallery closure need not mean that its holdings were destroyed. Equally, a continuing institutional name does not guarantee that every earlier object survives.
Separate physical location, administrative custody, ownership, public visibility and research availability. These can change at different times. A specimen transferred to a university might become less visible to casual visitors while remaining available to researchers. A borrowed object can occupy a prominent gallery without becoming part of the museum’s permanent holdings.
Singapore’s natural history collections provide a verified starting point for this distinction, but a full investigation should follow the particular collection rather than extrapolate from the institutional timeline. Use inventories before and after a transfer, correspondence about custody and later research records. Where documents count different units, reconcile them before concluding that objects vanished.
Reconcile a fictional transfer
Harbour dispatches 120 numbered entries to another institution. The receiving inventory lists 150 physical pieces because 30 entries each contain a separate base and object, while the other 90 contain one piece. The arithmetic is 30 times two plus 90, giving 150. The larger receiving total does not prove an unexplained acquisition during transport.
Now suppose only 148 pieces are recorded. That is a discrepancy requiring investigation, not immediate proof of theft. A base might be packed separately, the receiving record might be incomplete, or the sender’s description might be inaccurate. Preserve both records and identify the unresolved difference. The historian’s task is to explain the evidence, including its limits.
War, disaster and political change can alter museums profoundly, but specific claims about loss or rescue need specific sources. Avoid repeating dramatic stories simply because they fit an institution’s commemorative narrative. Identify who recorded the event, when, with what access to information, and whether independent evidence supports the account.
CHAPTER 12 OF 14 · Test public claims and change
12. Recognise changing authority and community participation
A museum’s authority comes partly from expertise and partly from institutional arrangements: who can acquire, classify, interpret, approve labels and decide what is displayed. These powers have histories. Studying them helps explain why a change in governance or consultation can affect the public account of the past even without a major new building.
Community participation should be examined through what participants could actually influence. Were people invited to supply memories, correct descriptions, select objects, write text or share decision-making? Those are different roles. A photograph of a consultation meeting establishes that a meeting occurred; it does not establish that participants could change the exhibition.
The 2022 ICOM definition’s emphasis on participation is useful evidence of a professional commitment at a particular time. Historical research then asks how a specific institution interpreted such commitments, what procedures changed and how the people involved assessed the result. Stated ideals and implemented arrangements belong in the same account, but should remain distinguishable.
Compare two invented consultation records
Harbour’s first project asks former workers to record memories after all labels are final. Its second project invites contributors to review draft descriptions and records several corrections to job titles and working practices. Both involve participation, but the second provides clearer evidence of influence over the interpretation. Explain that difference rather than simply labelling one project inclusive.
Disagreement need not mean the history has failed. Different contributors may remember an event differently or value different aspects of a collection. Record the basis of the disagreement and avoid manufacturing a single voice that nobody actually expressed. Where a label must be short, supporting material can explain the evidence and uncertainty in greater depth.
Respect also involves recognising that not all knowledge is appropriate for unrestricted display. Historical inquiry should examine the relevant relationships and stated conditions rather than assume that every object or account becomes public merely because a museum possesses a record. The practical and ethical arrangements depend on the case; avoid turning one institution’s solution into a universal rule.
CHAPTER 13 OF 14 · Transfer and investigate
13. Transfer the method to a digital exhibition
Imagine that Harbour creates an online exhibition of 300 photographs selected from 2,000 catalogued images. It presents them through searchable descriptions, a timeline and visitors’ comments. The interface feels different from a gallery, but the historical questions remain: who selected the material, what was omitted, how was it described and what could the audience actually discover?
The selected photographs represent 15 percent of the catalogued images: 300 divided by 2,000. They do not necessarily represent 15 percent of all events, communities or years covered by the archive. Selection may favour images with clear dates, reproduction permissions or sufficient resolution. Those practical reasons can shape the historical story visitors encounter.
Diagnose a misleading digital success claim
The invented exhibition records 10,000 page views and 2,000 image downloads. A report says “20 percent of visitors downloaded an image”. That conclusion does not follow from these figures. The ratio of download events to page-view events is 20 percent, but one person can create multiple events. Unique people require different evidence, and even that evidence may have limitations.
Now a catalogue correction changes a photograph’s date from 1920 to approximately 1930. The timeline should reflect the revision, and a record should explain the reason. Otherwise a future researcher may encounter two conflicting screenshots without knowing that the institution corrected its attribution. Digital interpretation has a version history just as printed catalogues have editions.
Preserve selection criteria, descriptive revisions, permissions and the interface’s organising structure. A folder of image files alone does not recreate the exhibition visitors saw. Search order, captions, grouping and prominent links shaped that experience. The historical object of study includes both the collection and the presentation built around it.
The transfer test is whether you can explain one difference and one continuity. A digital interface can make remote discovery possible, while still selecting and framing evidence. Wider technical reach does not prove universal access, complete representation or reliable interpretation. Those remain questions to investigate with appropriate evidence.
CHAPTER 14 OF 14 · Transfer and investigate
14. Build a practical research route and answer common questions
Choose one museum and one change small enough to investigate: an opening, a collection transfer, a revised gallery, a new catalogue category or a community project. Read the institution’s chronology first, then identify the records behind the relevant claim. Keep the scope narrow enough that you can explain how the change occurred rather than merely announce that it happened.
Create a source sheet recording creator, date, purpose, location and the claim each item supports. Distinguish an object’s production date from its acquisition date, a building’s completion from public opening, and a retrospective account from a contemporary record. Record uncertainty while researching, when its cause is still clear.
What should a beginner do during a museum visit?
Choose one case or display and identify its main question. Read three labels closely. Note what is stated about dates, makers and provenance, and what is marked as uncertain. Ask how the objects were selected and arranged. Afterwards, compare the gallery’s interpretation with a catalogue record or an earlier publication, where available. The exercise should produce a checkable observation, not a judgement based only on whether the room felt impressive.
Is museum history the same as museum studies?
They overlap. Museum studies examines institutions, interpretation, collections and public responsibilities across several approaches. Museum history concentrates on change over time: how particular institutions and practices came into being and developed. This guide connects with the existing Museum Studies root while keeping chronology, causation and historical evidence at the centre.
Which museum was the first?
The answer depends on the definition and scope: first public collection, university museum, national institution or purpose-built building are different claims. Avoid repeating an unqualified global first when the source establishes only a narrower category. Explain the criteria and acknowledge earlier or different collecting traditions relevant to the comparison.
Can an institution write its own history reliably?
Its records and expertise can make its account an excellent starting point. Read it critically, as you would another source. Follow references, distinguish documented events from celebratory language, and compare institutional claims with contemporary records and visitors’ or contributors’ experiences. An official account is evidence, not a substitute for asking how its claims are supported.
How do I know whether my explanation is complete?
A reader should be able to identify the change, its timing, the people and arrangements involved, the evidence for the explanation and its remaining limits. Include a plausible alternative explanation where the evidence permits more than one. If a number matters, state its unit and denominator. If an object matters, identify it precisely enough that another person can locate the same record.
For further study, use the linked Ashmolean, British Museum, Louvre, National Museum of Singapore and Lee Kong Chian Natural History Museum histories as starting points into particular institutions. Pair them with the specialist eduKate articles on everyday objects, natural history collections and catalogue reading. The result should be an account that helps someone understand how a museum made knowledge public, how that process changed, and what its surviving evidence can and cannot tell us.