More appropriate name for TaxonRelationshipAssertion class

We have adopted the TaxonRelationshipAssertion element from TCS, but have never really discussed the name of the class.

In the discussion around vernacular names (#47), the class was frequently mentioned and @jar remarked that it is not appropriate to call it "TaxonSomething", as they are relations between Taxonomic Name Usages.

It seems logical to call the class TaxonomicNameUsageRelationshipAssertions, but we restricted the type of relationships in this class to the horizontal relationships between taxon concepts (isCongruentTo, includes, isIncludedIn, overlaps, intersects and excludes) and there are several relationships between TNUs in the standard that are not in this class.

It would be nice to come up with a name for the class that reflects the types of relationships we allow in it. Any suggestions? I would prefer to not call it RCC-5 Relationships (if only because we have a term that is not in RCC-5) or Set Relationships.

Sorry, I seem to go on and on! You see why I try to stay out of these discussions - it is out of consideration to you all.

You certainly don't need to apologize to me! As I've made clear I find this exchange extremely useful. I have other commitments today, and also as a courtesy to others, I'll keep this short, mostly because...

I think we don't disagree even if it appears to you that we do.

In reading this post and previous ones, I fully agree! And I hope I didn't sound like I disagree substantively with you, because I most certainly don't!

I have read your entire most recent post agree with pretty-much all of it. I'll only add a few comments to help clarify my position on a couple of details, to determine if you agree on those details as well:

Certainly if the namestrings differ we are talking different usages;

Pedantically, not always. If "Aus bus" and "A. bus" both appear in the same treatment of the same reference, they are both [part of? properties of?] the same TNU instance, even though they are different namestrings. The same would apply to misspellings within the same treatment (e.g., Aus buus). I had originally modeled TNUs with a single "canonical" namestring, but after recent discussions in Canberra, I am increasingly of the mind that we should model TNU:NameString as 1:M.

So - the way I read it, this is a agreement.

Same here!

you seem to suggest that by "biological entity" you mean something more than just a 'set' of organism that 'that a competent taxonomist might circumscribe'

Oh no! I certainly didn't want to suggest that.

Sorry! My bad! It was the millions of years thing that made me unsure. Having read your post above we are in full agreement.

My weasel condition is that the group only has to potentially be circumscribed, not that it is

Understood! Makes perfect sense now.

Almost always, without a stated circumscription, you won't know which group I'm talking about, so it might appear that the circumscription and the group are inseparable, or that the circumscription creates the group; but neither is the case.

Agreed, but the problem is that a large amount (vast majority?) of taxonomic literature doesn't go far beyond simply listing a species name and implying that it refers to a specific circumscribed set of organisms. If taxonomists had been better about this, we could definitely drop the "Assertion" part of the term we're talking about. But because so many TNUs involve such few details about the circumscription boundaries, we're left to the best judgement of third parties (accordingTo Reference) to make assertions about what was meant in a circumscription sense. This isn't disagreement, it's just an effort to further articulate my thinking on this.

Suppose we agree that mosasaurs are a group. Did the group of mosasaurs exist in the Cretaceous? No sensible biologist would say that it didn't. Did they form a taxon in the Cretaceous? I don't know - maybe their description in the modern era somehow caused the group to become a taxon way back then - that would be odd but maybe it doesn't matter.

Just to clarify what I meant, certainly there are taxon circumscriptions for organisms that have not been alive for millions of years. But I would argue that the circumscription didn't exist until a taxonomist asserted it. The organisms just died and left progeny (or not) and fossils. So while the organisms existed millions of years ago, the taxon circimsctiptions were born int he minds of the taxonomists who studied their fossils. At least that's my take. Let me know if you disagree.

Had humans and scientific communication never evolved, would the group have been a taxon?

Well... only if some other life-form (cephalopods?) developed sufficient cognition to communicate abstract ideas with each other, and in doing so created circumscriptions of organisms that they found to convenient to refer to as discrete entities. But that's just my take....

Perhaps there would have been groups, but no taxa at all, in that situation.

Well, I would only call it a "group" if it had a boundary between it and other groups. I'm not sure we have many (any?) clear examples of unambiguous boundaries between groups of organisms in nature.

If we just disagree on our preferred usage of the word "taxon" that's fine, because I don't expect anyone in taxonomy to use it the way I personally like to use it. I won't push.

Same here -- and neither will I.

(Re existence, let me try an artificial analogy. You and I are standing in a room, and I am trying to talk to you about a circular area on the floor that I have "discovered" (there is something useful to be said about it). There are many such areas and at first you have no idea which such area I might be talking about. Then I take a marker and draw a circle on the floor, and tell you that the area I am talking about is the one inside the circle that I have drawn. Then you know what I am talking about.

Nice example! And to carry through with the metaphor, if two of us were in the room, and we were both pedantic (as taxonomists tend to be), one of us might assume you meant only the area within the marker line, and the other might assume you meant the area within the marker like and the area under the marker ink (i.e., the circumference on the inner boundary of the marker line, or the outer boundary). The problem in biological taxonomy is that (continuing with the metaphor) the area under the ink can sometimes be as large or even larger than the area circumscribed within (but not covered by) the ink.

The problem isn't about the circular areas -- it's with the methods we use to communicate the boundaries of the circle.

But that point aside, I think your example of the circle on the floor is a perfect description of the distinction between the "idea" of "thing" along with the way we describe the boundaries of that "thing", and the "thing" itself. And in that sense, we are in absolutely full agreement!

I've been going over this in my head to try to do justice to your point. When we are having an argument about a RCC-5 relationship, whether the argument is one of how to interpret the TNU sources or a scientific argument about evolution, what we care about is not the relationship per se, which has few interesting properties, but whether the relationship holds (exists) at all. (I think this agrees with what you said.)

Yes, it does!

It is fo some value to talk about a claim (e.g. that a relationship exists or holds) even if the claim is not asserted. So 'assertion' may be a red herring.

Ah! OK, this paragraph helps me understand where you're coming from. I don't disagree, but I need to think about it some more.

But having a table of 'assertions' suggests that the author of the source is making the assertion (who else would have done so?),

I'm not sure I follow -- can you elaborate on who you mean by the "author of the source". Sometimes one of the references associated with one of the TNUs explicitly makes the relationship, in which case the accordingTo Reference is the same as one of the TNU References. But in many (likely most?) cases, the accordingTo will be a separate Reference, not the same as either of the References associated with the two TNUs. I suspect we both agree on this, but I wanted to confirm.

I guess I'm saying that denormalization is the key idea that lets us reduce the number of classes and tables when entities are roughly in 1-1 correspondence.

Yes, I agree! And that's why the idea of a new class of entity representing the "nameless" circumscription is so appealing. It's just that the devil is in the details, and I have yet to see a successful way to do this that didn't end up creating as many problems (if not more) than simply talking about relationships among TNUs. I'm keeping a VERY open mind, though -- so I'm happy to see an example of this that would scale sufficiently.

Still not sure 'asserted' adds anything.

I'm a lot less sure of this than I used to be, and am certainly willing to abandon all advocacy for including the word "Assertion" as part of the term! Not because of debate fatigue, but because my perspective has genuinely changed as a result of this discussion.

The author is making a table, one per relationship.

By Author I assume you mean the accordingTo reference?

They probably claim the relationships hold, otherwise they wouldn't put them in the table, and a relationship that doesn't hold doesn't exist and isn't really a relationship (think unicorns).

I need to think more about what you mean by "hold" in this context.

Best not even try to say what kind of proposition it is, or what is intended in expressing it.

Agreed!

I'm just noodling here - please see this as an exploration, not me trying to bully anyone.

Likewise for my posts and associated ramblings!!!

I don't think the way we spell this class (the class of whatevers they are that the rows of the relationships table are 'about'), whether it contains 'Assertion' or 'Claim' or 'Relationship' etc., is the most important question we face. The documentation is much more important than the label.

I COMPLETELY agree! But this discussion still has been, and continues to be, very interesting and helpful for me!

I am allergic to the word "representing" because technical sources seem to all use it differently and/or at odds with ordinary language, and usually without any definition or analysis.

Got it, understood. Tough stuff to talk about, these things!

Thanks @jar398, I agree with @deepreef that your blog post was not at all aggravating. In fact I think it is very useful and very timely considering where we are in the documentation process.

Regarding the name for the class that is currently called 'TaxonRelationshipAssertions', I think it might be better if the name of the class reflects the kind of relationship rather than the kind of things at either end of the relationship. TCS used Set Relationship as a category of relationship types. I don't particularly like it for a name (as I think they might not quite be set relationships), but still prefer it over Taxon Relationship or TNU Relationship, because parentNameUsage and acceptedNameUsage are relationships between TNUs as well (and in TCS they were handled as Taxon Relationship Assertions)). I am happy now to lose the 'Assertion' bit from the name.

I like the formalisation (if that is the right word to use) of the relationship between TNUs and taxonomic groups and would not at all be unhappy to not use the term 'concept' anymore.

On 3/22/20 5:17 PM, Richard L. Pyle wrote:

Pedantically, not always. If "Aus bus" and "A. bus" both appear in the same treatment of the same reference, they are both [part of? properties of?] the same TNU instance, even though they are different namestrings. The same would apply to misspellings within the same treatment (e.g., Aus buus). I had originally modeled TNUs with a single "canonical" namestring, but after recent discussions in Canberra, I am increasingly of the mind that we should model TNU:NameString as 1:M. This seems OK, I defer. One just has to be careful not to impute equivalences that competent taxonomists might disagree about.

Just to clarify what I meant, certainly there are taxon circumscriptions for organisms that have not been alive for millions of years. But I would argue that the /circumscription/ didn't exist until a taxonomist asserted it. The organisms just died and left progeny (or not) and fossils. So while the organisms existed millions of years ago, the taxon circimsctiptions were born int he minds of the taxonomists who studied their fossils. At least that's my take. Let me know if you disagree.

Yes, circumscribing is something that people do (or maybe robots), and they do it with mixed success. But the group does not comport with people, it is a biological entity like a molecule or gene, and doesn't care at all whether it's circumscribed.

Perhaps there would have been groups, but no taxa at all, in that situation.

Well, I would only call it a "group" if it had a boundary between it and other groups. I'm not sure we have many (any?) clear examples of unambiguous boundaries between groups of organisms in nature.

Hmm. I usually think of a group as a kind of set, a set meeting certain qualifications (circumscribability). A set is equivalent (iso-ontic) to an enumeration of its members, or to a rule or procedure that lets you decide whether something before you is in it or not. If there is an issue of clarity, that has to do with human communication, which I think we've established doesn't bear on the group. But I don't think this is an important point, since we have no reason to talk about uncircumscribable groups - if someone said "set" was the right word I might grumble but accede.

I'd like to amend what I said, which was something like 'you may not know what group I'm talking about', to 'maybe neither of us knows what group we're talking about'. So we can happily talk about A. bus without nailing down exactly what either of us thinks it is (what is or isn't an A. bus in every case). This does not make groups or taxa defective or special in any way - this sort of thing is the rule, rather than the exception in biology, science, and human communication in general. Molecule, gene, protein complex, cell, organism... none of these is exact and at the boundary, so two communicating people may need to stop at any point and say, well what do you mean, where is the boundary? Do you consider the ions to be part of the channel, or not? So the 'circumscriptions' of these things can be elaborated as needed in order to answer questions people care about.

I know taxonomists get burned in special ways around this kind of 'vagueness' or incremental interpretation, and that's why they are so touchy about 'taxon' (and apparently even 'group'). I say, just get comfortable with the weirdness of philosophy of language and philosophy of science, and take it easy. This attitude is supported very nicely by the TNU idea.

(Re existence, let me try an artificial analogy. You and I are
standing in a room, and I am trying to talk to you about a
circular area on the floor that I have "discovered" (there is
something useful to be said about it). There are many such areas
and at first you have no idea which such area I might be talking
about. Then I take a marker and draw a circle on the floor, and
tell you that the area I am talking about is the one inside the
circle that I have drawn. Then you know what I am talking about.
Nice example! And to carry through with the metaphor, if two of us were in the room, and we were both pedantic (as taxonomists tend to be), one of us might assume you meant only the area /within/ the marker line, and the other might assume you meant the area within the marker like /and/ the area under the marker ink (i.e., the circumference on the inner boundary of the marker line, or the outer boundary). The problem in biological taxonomy is that (continuing with the metaphor) the area under the ink can sometimes be as large or even larger than the area circumscribed within (but not covered by) the ink.

Yes this is exactly the communication problem that we aspire to fix... not fix, but make better. Vagaries of interpretation are inherent in human communication, so we can only improve, not solve in any final way.

The problem isn't about the circular areas -- it's with the methods we use to communicate the boundaries of the circle.

Absolutely

But that point aside, I think your example of the circle on the floor is a perfect description of the distinction between the "idea" of "thing" along with the way we describe the boundaries of that "thing", and the "thing" itself. And in that sense, we are in absolutely full agreement!

yes, on first reading what you wrote I was worried, but now I agree that we agree (for now at least)... in the context of that, I am curious what you think of the words "group" and "taxon" now. I don't really like "group", I threw it out there for lack of an alternative. Maybe if "group" is defined as whatever it needs to be in order to serve in RCC-5 articulations, as a term of art, it could serve us?
But having a table of 'assertions' suggests that the author of the
source is making the assertion (who else would have done so?),
I'm not sure I follow -- can you elaborate on who you mean by the "author of the source".

Oh sorry. We are talking about a potential vocabulary (TNC?) to use in some data dump, probably a small collection of CSV or TSV files (although eventually we might aspire to RDF). That set of files is a "source" (information comes from it) and somebody was responsible for its creation and dissemination. That someone is who I mean by "author". Now this author may just be relaying selected claims made by other authors, but they still bear some responsibility, so I name them as the author. And some claims they make will be about entities designated by TNUs (or more precisely TNU-denoting expression), with the intent of referring to what name N means according to its usage in some other source S.

Sometimes one of the references associated with one of the TNUs explicitly makes the relationship, in which case the accordingTo Reference is the same as one of the TNU References. But in many (likely most?) cases, the accordingTo will be a separate Reference, not the same as either of the References associated with the two TNUs. I suspect we both agree on this, but I wanted to confirm.

yes
I guess I'm saying that denormalization is the key idea that lets
us reduce the number of classes and tables when entities are
roughly in 1-1 correspondence.
Yes, I agree! And that's why the idea of a new class of entity representing the "nameless" circumscription is so appealing. It's just that the devil is in the details, and I have yet to see a successful way to do this that didn't end up creating as many problems (if not more) than simply talking about relationships among TNUs. I'm keeping a VERY open mind, though -- so I'm happy to see an example of this that would scale sufficiently.

I see no reason to introduce any TNU-like things that are not TNUs, or identifiers for taxa or groups or collections of TNUs, but I really did not follow that part of this discussion and I'm sure I missed the motivation. The denormalization I meant has expressions (local identifiers maybe, or little groups of cell values) denoting TNUs, while the claim made by the row (or part of a row) is really about not the TNU, but about the group that is the usage of the name in the source. We do not introduce a name for the group, we at most introduce a local identifier for the TNU (although the TNU doesn't require a global identifier since it's easily identified by the reference string that identifies the source together with the namestring whose usage in the source we care about).

This need not remain the case. If I were trying to do a rich DL ontology I would certainly introduce group as a class, since an ontology is just a theory of a domain, and says nothing about how you write down your data, whether things have names, etc. That doesn't seem to be what we're doing here.

Still not sure 'asserted' adds anything.
I'm a lot less sure of this than I used to be, and am certainly willing to abandon all advocacy for including the word "Assertion" as part of the term! Not because of debate fatigue, but because my perspective has genuinely changed as a result of this discussion.
The author is making a table, one per relationship.
By Author I assume you mean the accordingTo reference?

no, see above, the author of the pile of stuff that contains the TNC-vocabulary-enriched tables whose rows claim that relationships hold.
They probably claim the relationships hold, otherwise they
wouldn't put them in the table, and a relationship that doesn't
hold doesn't exist and isn't really a relationship (think unicorns).
I need to think more about what you mean by "hold" in this context.

If R is a foo-relationship between X and Y, then R holds iff X is foo-related to Y. E.g. if M is the marriage between Alice and Bob, then M holds if Alice is [currently] married to Bob. Like "is in effect".

I had great hopes that we might come to see TaxonomicNameUsage as simply a factual collection of typed and annotated name/reference pairs. But I’m starting to worry that TNU sensu TNC are so swept up in taxon/concept thinking that we run the risk of producing yet another TCS and completely failing to deliver the simple and pragmatic, extensible standard required for reusable interchange of nomenclatural and taxonomic data.

At least for biodiversity informatics, data in the realm of taxonomic thinking always manifests as taxonomicNameUsage - contextually linked arrangements of names and name-references in a reference (s.l.). All such usages are facts. They are also strongly typed using +/- precise definitions mandated by the Codes of Nomenclature and convention. These facts provide everything we need for data interchange. There is no need to make stuff up.

At this level there is no such thing as taxon, taxonomic concept or circumscription. These are all abstractions of taxonomicNameUsage. None of them exist without it. If I want to make a concept I have to mint a TNU - according to the rules. If I build a nomenclatural or taxonomic system and show it to the world I am exposing TNU. If we need to share or consume data we do it using TNU. All of our systems have elements that can be deconstructed to this layer and reconstructed from it. Shared and reused they may become names or taxa, even concepts, but that’s not required of our standard.

Our primary task is to document the TNU vocabulary. If we can do this in a concise and unambiguous manner, adhering to our use cases, we will have the means for data interchange and reuse, first priority, and then perhaps, the foundation required by others for the axiomatic development of taxonomic abstraction using more formal methods.

Greg, that's a perfectly coherent position, and separating the nomenclature and reference vocabulary from anything biological such as RCC-5 makes perfect sense.

In a sense you and I are getting worked up about the same thing. If biology ('groups') is in the charter of the TNC interest group (as the name itself implies), we need to be very careful about how we deal with it, since it is known to be procedurally toxic; that was my aim since I just assumed the interest group had committed to dealing with RCC-5. If they are not in the charter, then we proceed exactly as you describe. What you are inspiring me to think is that it might help (I am not sure) to split the overall TNC effort into the nomenclature/bibliography/TNU part, the part I think of as a variety of library science, and the biological (RCC-5 etc.) part, where we're wondering what the words in question actually mean and what the biological world is like.

This could mean two separate vocabularies and possibly even two separate interest groups ('TN' and 'TC' instead of 'TNC'), although probably not since the same people care about both.

There is a call for a standard way to express biological claims, even ones that are hypothetical. Ultimately nomenclature has no purpose on its own, it is just there to serve this purpose. So we shouldn't say forget about biology entirely.

It is hard to separate them because the same people tend to do both, and rigorous discussion of biology depends on TNUs i.e. rigorous reference to written sources. Educating a larger community that the distinction is important is an uphill battle. But I think we have to try.

I also worry about RCC5 and TCS-style Assertions. If it were not for the fact that they are important for version management and notification services in our taxonomic overlays I would be inclined to suggest taking “assertions” out to a future extension using a broader task group. The bulk of the assertion discussion here would have had most of our intended TNC client base running for cover.

TCS uses TaxonRelationshipAssertions “… to express relationships between concepts without actually presenting new concepts. “ and comments, “In cases where this is occurring it may be worth considering whether the data source should actually be publishing their own concepts”. Such relationships can be polymerous (2+). The preferred option being to use taxonConceptRelations (from/to), which allow for all the kinds of relationships that can occur between TNU (synonymy, parentage, RCC5-style, … etc).

In taxonomic works, it is common to find TNU accompanied by RCC5-like annotations which are clearly within-TNU relationships that could be expressed as relationship TNU. Taxonomic synonyms and misapplications are for the most part unambiguous RCC5-style assertions. Objective same-name, same-type and same-concept relationships on the the other hand are often hidden in text where the absence of a formal vocabulary or convention can make them difficult to discover let alone assert. Such relationships are often complex, being both nomenclatural and mereological, and perhaps showing dissimilar ostensive and intensional patterns that may themselves have RCC5 relationships with additional “taxa”. I won’t mention misapplications.

In the our model, to allow for simple capture of the actual relationships within a work, we generalise all relationships using types (within the TNU class) and decompose assertions to TNU.

It would just be great if we could share them.

I'm not sure I understand all of that but it inspired one observation I hadn't made before, which is that the provenance of an RCC-5 claim or 'assertion' is independent of that of the two TNUs it references (which are not claims at all, but designators). One typical use case - I think the one you give - is where two TNUs refer to the same source, and that source states an RCC-5 claim relating their intended groups. That would be typical when we transcribe information in a single taxonomic treatment. Another typical case is where the two TNUs come from different sources and some third party (source) is suggesting an 'alignment' relating the TNUs in the two sources - this is Nico Franz's use case. A third, possible but atypical case, is that the two TNUs come from the same source, but an independent author (a different source, perhaps the source in which the claim is written) is suggesting a relationship between their groups that is not stated in the first source, perhaps even disagreeing with that source (i.e. saying that the source is internally inconsistent). In spreadsheet terms, to accommodate all of these cases, a TaxonRelationshipAssertion table requires an 'according to' column for the source of the claim (although you can imagine the column being elided if there is some systematic way in a particular spreadsheet to infer it; e.g. there is only one source and all claims are from the author of that source; or the claims form an 'alignment' and come from the author of the table itself, not any of the sources of the TNUs).

I don't think you can get away without saying something about what an RCC-5 relationship means, i.e. how we are all supposed to decide whether it is correct to claim (say, write, publish) it or not. That is why Rich and I dove into a discussion of 'groups' - you probably need the idea to guide people using the new vocabulary in the correct use of the vocabulary.

If we can avoid relaying the kind of discussion that Rich and I just had, I agree that's best. There are other ways to get the point across, e.g. by example. We're not going to prevent people from creating more nature/nomenclature quagmires by talking about them, but we can steer people the right way by using language carefully ourselves. Part of that is erecting a rhetorical firewall between TNUs (human sphere) and groups (nature).

(sorry that should be 'get away without saying' not 'get away with saying')

Lots to get caught up on... and don't want to add to much more. But some thoughts:

@jar398 : I found my head nodding in agreement (sometimes vigorously) in reading most of what you wrote. There may still be a small disconnect on the idea that groups exist as biological entities independent of a human's or robot's perception of them, but that's a philosophical debate that will likely remain unresolved among taxonomists for a long time, and is largely outside the scope of this TNC discussion (though I think it's extremely helpful to understand as background for this discussion).

On the terminology, I think I still prefer "taxon" to "group". There are trade-offs between the respective confusion caused by using a well-known term that means different things to different people, vs. using a less-familiar term that requires people to spend time thinking about and discussing to get their heads around. I went through this same issue when deciding whether to use "Protonym", or adopt (and slightly modify) "Basionym". In that case, I felt that using "Basionym" would have required changing the fundamental meaning of the term too much, whereas in this case, I think our use of "Taxon" falls within the scope of current usage, it just needs to be qualified and defined well, within our context. In RCC-5 terms. Basionym overlaps with Protonym, but "Taxon" in our sense is included within "Taxon" sensu lato.

If R is a foo-relationship between X and Y, then R holds iff X is foo-related to Y. E.g. if M is the marriage between Alice and Bob, then M holds if Alice is [currently] married to Bob. Like "is in effect".

OK, that helps. I'm not sure this applies here, though, because the area under the ink vs. the area circumscribed by the ink is FAR smaller for something like a marriage, vs. something like a taxon circumscription -- mostly due to the crude ability to communicate where the ink actually is laid down in the latter. When the ratio of ink:circumscribed area is very small, then we approach a situation of objectively discernible entities. But when the ratio is higher (sometimes > 1), then there is not solid basis for determining whether a claimed relationship holds (or not).

@ghwhitbread :

I had great hopes that we might come to see TaxonomicNameUsage as simply a factual collection of typed and annotated name/reference pairs. But I’m starting to worry that TNU sensu TNC are so swept up in taxon/concept thinking that we run the risk of producing yet another TCS and completely failing to deliver the simple and pragmatic, extensible standard required for reusable interchange of nomenclatural and taxonomic data.

I have the exact same hopes, and I actually think this discussion bolsters those hopes. Why? Because I think the fact that we're now focused on hammering out long-standing subtle issues with the higher abstraction of "TaxonRelation", it suggests that we're very close to full consensus on TNUs as simple, factual collections of typed and annotated name/reference pairs. I don't see any of the discussion in this thread altering the definition of a TNU in any way -- I see it as merely clarifying what they are (or, using my preferred words, are intended to represent).

Our primary task is to document the TNU vocabulary. If we can do this in a concise and unambiguous manner, adhering to our use cases, we will have the means for data interchange and reuse, first priority, and then perhaps, the foundation required by others for the axiomatic development of taxonomic abstraction using more formal methods.

Agreed, but I think this thread has helped me clarify in my own mind what TNUs are, and how they serve as much more than a simple index of name-reference pairs. I think this thread has informed our effort to document the TNU vocabulary, in the same sense that any use-case would. While I certainly agree that defining the TaxonRelation class is secondary to defining TNUs and their properties, it's still important to keep in mind. Having said that, I think perhaps a more critical topic of discussion going forward is to define a "Reference" class and its associated properties.

I agree with @ghwhitbread and @jar398 that it may be best to separate the Relationship stuff from the TNU stuff, at least until we nail down the TNU stuff sufficiently. And by "TNU stuff", I'm also including Reference stuff -- which should be part of our conversation later today.

I have separated off the TNU stuff into a separate issue (#48).

Regarding the name of the Taxon Relationship Assertion class, I think that there is now agreement that at least the 'Taxon Relationship' bit is okay. Rests us to decide whether to keep the 'Assertion' bit or not. I would like to make the case to keep 'Assertion' in the name of the class.

Like TCS, we have got taxon relationships in the Taxon Relationship Assertion as well as in the TNU (Taxon Concept in TCS) class. Compared to TCS, we have divided the taxon relationship types into relationship types that can be in TNU and types that can only be in Taxon Relationship Assertion. For the relationship type that can be in TNU, parentNameUsage, the accordingTo of the subject TNU – and mostly also that of the object TNU – is the same as that of the TNU, while for the RCC-5-type relationships that we keep in Taxon Relationship Assertion, both subject and object TNUs can have a different accordingTo as the Taxon Relationship Assertion and at least one of them has to.

I think it is the accordingTo that makes the object an assertion. Taxon relationships within the TNU are as much assertions as taxon relationships outside, but a TNU is already an assertion, so so is everything inside it. We should make that clear for the other taxon relationships as well.

tdwg / tnc

More appropriate name for TaxonRelationshipAssertion class #48