Why does this adjective produce a runtime problem?

Authoring Inform 7
1

[code]A multitude is a kind of backdrop. A multitude has a thing called the specimen.

To decide which object is the collection of (the singleton - a thing):
Decide on a random multitude that has a specimen of the singleton.

Definition: a thing is multitudinous rather than unique if the collection of it is a multitude.

The Quarry is a room. Gravel is a multitude. Gravel is in the Quarry.

A rock is a thing. The specimen of Gravel is a rock.

When play begins:
say “[The list of multitudinous things] is multitudinous.”[/code]
This code produces the following error:

I thought that since my description names “a random multitude” that “yourself” would never be tested. But it seems that it is. Do I have to loop over multitudinous things, or is there a way to make “a random” work here?

Curiously, when I changed “it” to “the item described,” I got a different error:

2

I’m not 100% sure what goes wrong in your code, but the logic is very suspect. You define that a thing X is multitudinous if and only if its collection is a multitude; but also by definition, the collection of X is a random multitude of which X is the specimen. So if X has a collection, X must be multitudinous. But X might have no collection, in which case it is undefined whether X is multitudinous or unique… it makes my head hurt.

Here is some code which achieves what I believe to be your purpose in much cleaner way:

[code]A multitude is a kind of backdrop.

Specimenship relates various things to one multitude.

The verb to exemplify (he exemplifies, they exemplify, he exemplified, it exemplified, he is exemplifying) implies the specimenship relation.

The Quarry is a room. Gravel is a multitude. Gravel is in the Quarry.

A rock is a thing. The rock exemplifies gravel.

When play begins:
say “[The list of things which exemplify something] is multitudinous.”[/code]

3

Thanks! That looks good, though it took me a while to sort out the transition from property to relation. I’m not a fan of relation syntax…

It also solves the next problem I was going to ask about, though. I wanted to say:

Understand the specimen property as describing a multitude.

But that wasn’t allowed. Fortunately, this is:

Understand "[something related by reversed specimenship ]" as a multitude.
4

Diagnosing the original error:

To decide which object is the collection of (the singleton - a thing):
   Decide on a random multitude that has a specimen of the singleton.

The “has” here is being interpreted as ownership, not the specimen property. That’s why you see the “parent of” error – it’s defining the collection as an object that contains the singleton’s specimen.

It’s possible to select an item by its property, but the relation is a better plan, over all.