Topic on User talk:Pasleim

"human" is not '''about''' a particular group. That's right.

But "human" '''is''' a particular group. It's the biggest possible group of humans. It's the all-set. --Eulenspiegel1 (talk) 20:44, 28 February 2019 (UTC)

Q5 is a class and not a group or set.

1. What's the difference between class and group?
2. human (Q5) instance of group of organisms known by one particular common name (Q55983715) subclass of group of living things (Q16334298) subclass of group (Q16887380). Thus, human (Q5) instance of group (Q16887380)

I was bold and changed group of living things (Q16334298) to living organism class (Q21871294).

The problem is, that "class" can be misunderstood as taxonomic rank. Especially, if we look at organism, many people understand "class" as a taxonomic rank. If we call it "goup", it is clear, that no taxonomic rank is meant. "Pisces" is a class of organism known by the particular common name "fish".

Also "group" is not wrong as every class is also a group. See https://www.w3.org/TR/rdf-schema/#ch_classes.

Furthermore, it makes no sense, to make two separate lines. What is the difference between "European" and "human"? Why is one an instance of "living thing group" and the other an instance of "living thing class"? Or what is the difference between "human" and "biota": They should be an instance of the same article, because "human" and "biota" together are "organism".

It doesn't make sense to differ: There are two possibilities:

1. We have an article with an individual: This is an instance of organism.

2. We have an article with several individuals: This is a subclass of organism and an instance of group of living things.

If you say, this is not enough and we need a third part, than explain the difference between class and group, please. When do we use "group" and when do we use "class"?

It is not easy to mix Wikidata instance class (Q21522908) with class (Q37517). It is more easy to mix group (Q16887380) with group (Q83478)

In a mathematical topic I would aggree. There it would be a problem to mix a group with the mathematical group.

Yet, in a biological topic this isn't a problem. Here it is a problem to mix a class (Q16889133) or a class (Q5127848) with the biological class (Q37517).

"European" is a class, "Europeans" is a group of humans.

A class is an abstract object, defined by properties all its instances share. A group of humans is a concrete object, you can take a group picture of each group of humans (in theory at least). In terms of RDF, "Europeans" is a set, or the class extension of the class "European". All members of the set "Europeans" are instances of the class "European".

While this sounds that we have to duplicate a lot, in reality either the class item or the group item is enough most of the time. For example both of the below versions are valid,

• Freddie Mercury is instance of the class "member of the band Queen",
• Freddie Mercury is part of the group "Queen"

but we agreed that we only use the second version.

Another example:

• Oslo is instance of the class "municipality of Norway",
• Oslo is part of the group "all municipalites in Norway"

Here only the first variant is used on Wikidata.

First of all: https://www.w3.org/TR/rdf-schema/#ch_classes describes it very well.

Classes are groups!

Now the answer:

"Europeans" is a class. Every instance of this class is a European. (The set of all its instances is the class extension of "Europeans".)

"Human" is a class. Every instance of this class is a human. (The set of all its instances is the class extension of "human".)

"Municipality of Norway" is a class. Every instance of this class is a Municipality of Norway. (The set of all its instances is the class extension of "Municipality of Norway".)

It's normal, that many classes are named after their instances. It's only a name convention whether you use the plural or singular. It doesn't change the class.

Your picture sentence I don't understand: You can make a picture of all Europeans (at least theoretical), thus "Europeans" is a group?

Yet, you can also do a picture of all humans (at least theoretical), thus "Human" is a group.

Yes, there is a different logic between "Europeans" and "Queen". Thus, I would understand if there would be a difference between "Europeans" and "Queen".

But there is the same logic between "Europeans", "human", "dog" etc.: In all of these classes the organism are not part of the class, the organism are instances of the class.

Groups are also very often abstract objects not created by logic. (See for example en:Ingroups_and_outgroups.) Also all classes are groups. See https://www.w3.org/TR/rdf-schema/#ch_classes. Maybe some groups are concrete objects. Yet, the most groups are abstract objects.

Groups can be abstract or concrete, I agree, any we should not mix them up. That's why I used above the example with the picture: "Europeans" is a concrete group, you can take a picture of this group but you cannot take a picture of an abstract group.

With the statement "Resources may be divided into groups called classes" in https://www.w3.org/TR/rdf-schema/#ch_classes I strongly assume that abstract groups are meant.

Now group of humans (Q16334295) is the class of all concrete groups of humans, while Q5 is a class or an abstract group.

You cannot make a photo of "Europeans", but you can make a photo of all instances of "Europeans".

Also you cannot make a photo of "human", but you can make a photo of all instances of "human".

Also you cannot make a photo of "Queen", but you can make a photo of all members of "Queen".

Also "Europeans" and "human" is less abstract because it is defined about their instances: If you know all about their instances, you know all about the class.

"Queen" is more abstract: There exists not only humans as part of the class. There exists also abstract things like inception (P571) or genre (P136).

"Europeans" and "human" don't have such abstract properties.

1. I think https://www.w3.org/TR/rdf-schema/#ch_classes describes it the best.

2. I don't like at all group of organisms known by one particular common name (Q55983715) but it is part of the bigger problem that taxa and taxon names are intermingled here on Wikidata. In addition this item has a different meaning in German and English which makes it even worse.

1. At https://www.w3.org/TR/rdf-schema/#ch_classes the first sentence is: "Resources may be divided into groups called classes." Thus, every class is a group. There exist many different groups. And a special kind of groups are called classes.
2. I have adapted the german labeling of group of organisms known by one particular common name (Q55983715). The german label described a subclass of common name (Q502895) and not a subclass of group of living things (Q16334298). Thus, the german label was wrong. Now, the english and the german label fit together. Nevertheless, it shows, that human (Q5) is an instance of group (Q16887380).
3. taxon (Q16521) is a subclass of class (Q16889133) . Yet, the description of taxon is: "group of one or more organism(s), (...)". This shows that taxons are classes AND groups.

1. You are claiming that Q5 is the all-set. So please read the relevant description for the difference between class and set.
2. Why not renaiming it to "class of organisms known by one particular common name"?
3. Q5 is not a taxon.

1. I had claimed that it is the all-set. This was a mistake. After reading the description I claim that it is a class.
2. When you can rename an article, than the two names are the same. If "class" and "group" are the same, than you can simple rename the article. If "class" and "group" are two different things, than ""group of organisms known by one particular common name" and "class of organisms known by ohne particular name" are two different articles. You cannot simple rename them. If you could rename them, than you could also rename "group of humans" into "class of humans".
3. Q5 is not a taxon. That's right. But taxon (Q16521) shows that something can be a group AND a class.

@Eulenspiegel1: may be you are looking for humanity (Q1156970)?

No, humanity is not a group of humans. A group of humans contains humans.

You can say "This person is a human." or "This person is a European." But you cannot say "This person is a humanity." You can only say "This person is part of humanity."

What would you say about a new class "group of humans known by one particular common name"? This new class would be subclass of "group of organisms known by one particular common name" and subclass of "group of humans".

I don't like the idea of having common subclass of "group of organisms known by one particular common name" and "group of humans". Because imho instances of the first class are classes and of the second are groups. I see the following fundamental difference between group and class here: group is a finite definite set of elements ("we can make a photo of a group") while class is an open indefinite ... (can't find a word, not "set" definitely) of elements (content of a class is dependant not only on time but also on some other qualifiers and ranks)

Yes, you cannot say "This person is a humanity." You can only say "This person is part of humanity." That's why "humanity" is a group and not a class.

Every class is a group. See https://www.w3.org/TR/rdf-schema/#ch_classes.

Also there exist definite sets (e.g. "member of Queens" ) and indefinite sets (e.g. "Europeans"). I would say that "human" is much more definite than "Europeans". Nevertheless, it doesn't matter if "Human" is as definite as "member of Queens" or as indefinite as "Europeans" or something between. In all this cases it is a group of humans.

Or see Bach family (Q169509). It's completely indefinite who belongs to it and who not: If we go more generations backwards, then more people belong to the bach family. If we go long enough backwards, than all people belong to the same tribe. Thus, all humans would belong to the bach family.

You cannot say "This cell is a human." You can only say "This cell ist part of a human." With your logic "human" is a group (of cells).

Humanity relates to human as human relates to cell: Human is not a "group of cells" and Humanity is not a "group of humans".

Pasleim, do you agree, that "human" is as abstract as "Europeans"? In both cases you cannot make a photo of the class itself but of the instances of the class.

@Infovarius:Do you agree, that it doesn't matter whether a class is as definite as "member of Queen" or as indefinite as "Europeans" or "Bach family"? In all this cases it is a group of humans.

Okay, back up a bit.

Unfortunately, the terms "group" and "class" have taken on specific meanings in Wikidata context that don't match up perfectly with their regular uses. Classes are things that can be targets of subclass of (P279) and instance of (P31), groups are collective objects that things can be part of (P361). "Collection" has domain-specific meaning, a "series" is an ordered "group" that can be targeted by part of the series (P179) with appropriate qualifiers, and "set" is avoided entirely except in referencing particular mathematical concepts. There's also uncodified label sort-of standards which I think Pasleim was referencing above regarding the difference between "European" and "Europeans" and how that relates to the class-group distinction.

Certain properties are handled differently for classes than for other items. Frequently, class X PY Z means that (each instance of X) PY Z. Taking a fictional example, if there are these two 1kg boxes, which shall each be called a "foo", the item "foos" (referring to the two boxes as a "group") could have mass 2kg, whereas the item for the "class" "foo" would have mass 1kg, referring to the mass of each instance.

None of this has been written into the glossary, or afaict documented anywhere, because we are all very bad at this. :P

Right. And human is a group because of your part of argument:

human (Q5) part of humanity (Q1156970).

Or the other way around: human brain (Q492038) part of human (Q5).

Of course, you have to differentiate between the qualities of a class and the qualitites of the instances of the class.

human brain (Q492038) part of human (Q5) means that each instance of human brain (Q492038) is part of an instance of human (Q5). human (Q5) part of humanity (Q1156970) means that each instance of human (Q5) is part of humanity (Q1156970). part of (P361), when used in an item for a class, refers to each instance of the class, as it is with many other properties.

Only classes can be targets of instance of (P31). 4984492 items have instance of (P31) human (Q5), requiring that Q5 be a class.

It seems, you think, a ressource is either a group or a class. That's wrong. See https://www.w3.org/TR/rdf-schema/#ch_classes:

''"Resources may be divided into groups called classes."''

Thus, every class is a group.

As I explained in my first comment, those terms "class" and "group" are used in a specific manner on Wikidata, which may not match up with how others (such as the W3's RDF standard) use them. On Wikidata, classes are valid targets of P31/P279. Things which are not classes are not valid targets of P31/P279.

1. You said, it's not written in the glossary. How do you know, that WIkidata is not W3C conform?

2. Why is Wikidata not W3C conform? What is the benefit of misregarding the W3C RDF standard?