Why half the world’s museum specimens are wrongly labelled

My colleagues who work in museum documentation have rightly been indignant about an article published by the Daily Telegraph yesterday, under the attention-seeking headline ‘Half of world’s museum specimens are wrongly labelled’.1

It’s easy to see why: Knapton’s and the Telegraph sub-editor’s more striking assertions don’t seem to reflect the more careful conclusions in the original article by Zoë A. Goodwin, David J. Harris, Denis Filer, John R. I. Wood and Robert W. Scotland on which she based her report.2

Let’s look first at the original research. Goodwin and her colleagues rightly note the importance of museum collections for research into subjects such as species distribution. The data about these collections is increasingly being aggregated into global digital resources. The research looked at the labelling of 4,500 specimens of African gingers (the genus Aframomum) held in 40 herbaria in 20 countries. (Note that the research was based on stored specimens: we’re talking about the tags attached to specimens in drawers here, not the labels alongside objects on public display.) Compared with the most recent systematic survey of the genus, at least 58% of specimens had been given a ‘wrong name’. The authors also noted a significant recent increase in the number of specimens held in collections, with over 50% of those examined having been collected since 1970. These conclusions were cross-checked against specimens of the family Dipterocarpaceae: it was found that 29% of specimens collected together but sent to different herbaria had been given different names in the different collections. The authors suspect that these findings may also be true of collections of temperate species; and also  that the situation may be even worse for entomological collections.

But what is meant by ‘wrong name’? The 58% of names that were ‘wrong’ included specimens that were identified only to the family or genus (correct but imprecise), or which had been identified using a synonym – a form of the species’ name that was acceptable at some point, albeit not the form currently in use. So, when looking at the 560 names attached to 49,500 specimens from the genus Ipomoea in the Global Diversity Information Facility, it transpired that only 16% of names were invalid, erroneous or unrecognised; 11% were for genera or higher taxonomic levels; and 40% were (previously-)acceptable synonyms. So the Telegraph’s headline should perhaps have been ‘Half of the specimens in tropical plant herbaria have been incorrectly or vaguely identified, or given an old-fashioned name’.

Nonetheless, the situation is still worrying. How did we get here? Three possible causes struck me immediately:

First, poor handwriting. Collection data is often transcribed from correspondence or field notes; and not everyone’s handwriting is entirely legible. Unless the person doing the transcribing knows exactly what to expect, errors will appear. The same is true when transcribing handwritten collection data into a digital repository.

Second, bulk accessions, when a collection is listed under a summary entry because there isn’t time to list all the specimens individually straight away. Sometimes, it can be a very long time before a collection is properly processed, identified, listed and labelled. If the original collector is unavailable, or the person doing the listing is not sufficiently knowledgeable to be quite sure what they’re dealing with, errors or imprecision will creep in.

Third, lack of resource. Specifically, natural history classifications change far more often than non-scientists like me would expect. It’s therefore nigh-on impossible for curators and documentation staff to keep checking all their specimens  to make sure they’ve not been reclassified and/or renamed by another researcher. Significantly, the article compared all identifications which individual specimens carried at any point since they were collected, with the most recent research on the genus – so recent it’s still in the press. And as the authors point out,

Even when an economically important group such as the Dipterocarps has been recently revised…, this knowledge is not necessarily transferred to accurate names in herbaria.

Goodwin and her colleagues, too, have an explanation for the state of affairs, and I think they bear me out:

First … too few taxonomic revisions across entire geographical distribution of taxa in recent times means that the taxonomy and nomenclature of these groups are provisional and many specimens remain wrongly named, unrecognised and/or not determined for decades….

Second, the number of available specimens for any sizeable group is considerable….

Third, the number of herbaria has increased greatly, which means that there are too many herbaria for a given expert to visit or request loans from. Rapidly increasing numbers of specimens in increasing numbers of herbaria are not being revised because there are too few taxonomists.

Similarly, I wouldn’t disagree entirely with Mark Carnall’s not-entirely-serious list of excuses for why we might have reached this state of affairs (although, as an art historian, I’d say he’s obviously wrong about no. 6 – though I’ll give him the archaeologists, of course). In particular,

It’s the loss of subject specialist knowledge in a hard hit museum sector where we are losing expertise year on year.


We’re only half way there. When we started, none of them were correct. Give us another 300 years and we’ll finish the job.

In other words, there’s a great backlog of work to do to get this information in a fit state, and we’re losing the expertise required to attack it. I spoke about the historic reasons why there may be problems in current museum documentation in my recent lecture to the Collections Trust annual conference, and I’m pleased that this research brings the problem to more general attention. Hopefully, this will help us as we argue for greater investment in museum documentation.

But I’ll end on a happier note, and give due credit to Klaus Taschwer, whose article reporting the same research in Der Standard provides a more measured account of the research – and links through to the article itself, and to similar studies.3


  1. Sarah Knapton, ‘Half of world’s museum specimens are wrongly labelled, Oxford University finds’, The Telegraph (published online 17 November 2015) <http://www.telegraph.co.uk/news/science/science-news/11998634/Half-of-worlds-museum-specimens-are-wrongly-labelled-Oxford-University-finds.html> accessed 17 September 2015. []
  2. Zoë A. Goodwin et al., ‘Widespread mistaken identity in tropical plant collections’, Current Biology, 25/22 (16 November 2015), R1066-R1067; DOI: 10.1016/j.cub.2015.10.002 []
  3. Klaus Taschwer, ‘In Museen ist verblüffend vieles falsch bezeichnet’, derStandard.at (published online 16 November 2015) <http://derstandard.at/2000025852341/In-Museen-ist-verblueffend-viel-falsch-bezeichnet> accessed 17 November 2015. []

Leave a Reply