Here's a page-by-page breakdown of this chapter on Abstract Entities, explained in plain English.
Page 1 (Chapter 2 opening)
This page introduces what the whole chapter will cover. The chapter is going to use a method
from the last chapter — the "Quinean method" — to figure out whether abstract things (like
numbers or properties) really exist.
Learning points listed:
• Use Quine's method to check if abstract things exist
• Explain the difference between "abstract" and "concrete" things
• Look at an argument called "One Over Many" for why universals (like "roundness")
might exist
• Introduce "truthmaker theory" — a different way of figuring out what exists
• Introduce the "indispensability argument" — an argument that math must be real because
science needs it
The main text explains the Quinean method again simply: To figure out what exists, you (1)
decide what sentences you believe are true, (2) translate them into logic, and (3) see what things
have to exist for those sentences to be true.
Then it sets up the chapter's big question: if we already accept that ordinary physical things
(tables, rivers, stars, people) exist, do we also have to accept that abstract things exist — things
you can't see, touch, or locate anywhere?
Page 2 (page 61)
This page gives the motivating example: math. We say things like "144 has a square root" or
"some fractions are less than one." If these statements are true, and we use Quine's method, it
looks like we're committed to numbers existing. But numbers don't have a location, shape, or
size — they don't seem "concrete" like tables do. So are we stuck believing weird invisible
objects exist? Or do we have to say math is just false?
The author says: we'll deal with numbers later. First, let's talk about a similar issue with
properties — things like shapes, colors, and virtues (honesty, wisdom). If tables have
"properties" like shape and color, are those properties themselves separate abstract things
floating out there? The page ends by promising to define "abstract" more precisely.
Box 2.1 (Propositions): A side note explaining that "propositions" are another type of abstract
thing — the meaning behind a sentence, not the sentence itself. Two different sentences in two
different languages can express the exact same proposition.
Page 3 (page 62)
This continues the propositions box: "Everyone is in a good mood today" (English) and its
Italian translation are different sentences, but they mean the same thing — that's the proposition.
Sentences themselves might be concrete (ink on paper, sound waves), but the meaning they carry
is treated as abstract.
Then: The Abstract/Concrete Distinction section begins. It gives the first (traditional) way
philosophers try to separate abstract from concrete:
• Concrete things have shape, size, color, occupy space, and can causally affect things (you
can bump into a table).
• Abstract things (like numbers) supposedly lack all of that — no shape, no location, no
ability to cause anything.
But then the author warns: this neat definition starts breaking down once you look at trickier
examples (introduced on the next page).
Page 4 (page 63)
This page has two tables showing the problem.
Table 2.1 confirms the simple cases: tables, rivers, and stars all have shape, size, color, location,
and can affect other things — so they're clearly concrete. Numbers and virtues have none of
those features — so they're clearly abstract.
Table 2.2 shows trickier cases that don't fit neatly:
• Electrons: no shape or color, but do have size, location, and can affect things — so are
they concrete or abstract?
• Space, time, space-time: unclear on shape, no color, no location (space doesn't sit
"somewhere"), might affect things — very confusing.
• Colors like "red": no shape/size, but arguably has "color" in some sense, and whether it
occupies space depends on your theory.
• Individual colors of specific objects (like the redness of this particular apple): seems to
have shape, size, color, location — closer to concrete.
The point: the simple checklist approach fails once you push on hard cases. Most philosophers
still call electrons and space-time "concrete" and colors "abstract" — but not because they fit the
checklist consistently.
Page 5 (page 64)
Since the checklist approach failed, this page offers a better way to define "abstract": abstraction
is the process of stripping away features of an object until only one feature is left.
Example: Take a table. Imagine mentally removing its height, shape, and material, leaving only
its color. That leftover color is an "abstraction" from the table.
Second example (numbers): Japan has 6,852 islands. If you ignore every individual detail about
each island (their shapes, sizes, colors) and just focus on how many there are, you're left with the
number 6,852. That number is an abstraction from the concrete islands.
Key definition given in the margin: Abstraction has two senses:
1. Psychological — a mental act of thinking about one feature while ignoring others.
2. Metaphysical — an actual relationship between two things, where one lacks features the
other has.
The book says it will use the metaphysical sense going forward: abstract objects = things that
are abstractions from other (concrete) objects.
Exercise 2.1 then asks you to classify seven things as abstract or concrete: beauty, the Big Bang,
the Earth's core, dreams, electromagnetic fields, Romeo and Juliet (fictional characters), and
God. This is meant to make you apply the definition to hard borderline cases.
Page 6 (page 65)
New section: Universals and the One Over Many Argument.
This raises the question: do properties like "roundness" or "negative charge" exist as separate
abstract things, on top of the round objects and charged particles themselves?
Key idea — Universals: A universal is something that can be "instantiated" (had by, or present
in) many different things at the same time, in different places. Roundness is the example: a
coffee mug, a doorknob, and an iPhone button can all be round at once, in different places. If
roundness is a real thing, it must be a universal — because it's wholly present in every round
object simultaneously (not split into pieces the way your body is split between a chair and a
desk).
Contrast — Particulars: things that can only exist in one place — not repeatable. (Example
from Aristotle: "man" is universal, but "Callias" — a specific person — is a particular.)
Plato and Aristotle both believed in universals, but understood them differently. Plato called
them "Forms" — perfect, ideal versions of things like Beauty, Justice, and Goodness. Plato
thought the Forms had three special features:
(i) They're transcendent — outside space and time entirely.
(ii) They're ideal — perfect. The Form of Justice is completely just, with zero unjust in it.
(iii) (continues onto next page) — they can only be known through reasoning, not the senses.
Page 7 (page 66)
This finishes explaining Plato's third point about the Forms: real-world things (art, laws, people)
that display beauty or justice always have some imperfection mixed in — some ugliness or
unfairness. Only the pure Form itself is 100% perfect. And Plato held that these Forms can only
be understood using pure thought/reason — never through your eyes, ears, or other senses.
Box 2.2 — A Priori vs. A Posteriori knowledge: This box explains a key philosophical
distinction that connects to Plato's claim:
• A priori = knowing something without needing observation — just by thinking it
through. Example: knowing "my brother is a bachelor" means "my brother is unmarried"
— you don't need to go check, you just understand the words.
• A posteriori (also called empirical) = knowing something through observation or
experience. Example: knowing the color of the Nile River — you'd have to see it
yourself, see a photo, or hear someone describe it.
• Empiricism = the broader view that all our knowledge ultimately comes from
experience.
The page ends by tying this back to Plato: he thought you can observe beautiful things or just
institutions, but you can never directly observe Beauty or Justice themselves — those Forms can
only be grasped through the intellect, a priori.
Continuing page by page.
Page 8 (page 67)
This page finishes comparing Plato and Aristotle, then introduces key vocabulary and the "One
Over Many" argument.
Finishing the Plato point: What we see with our eyes is only a "shadow" of the real Form. To
truly understand the Form itself (which is perfect), you need pure reasoning, not your eyes.
Aristotle's different view: Aristotle disagreed with Plato about where universals exist. Instead
of putting them in some separate transcendent realm, Aristotle said universals are immanent —
meaning they exist right inside the ordinary objects that have them, located in space and time just
like those objects. So "roundness" isn't off in some other realm; it's right there in the round mug.
David Armstrong is mentioned as a modern philosopher who agrees with Aristotle's approach.
Armstrong thinks properties are: abstract, repeatable (universal), located in space/time (fully
present wherever they're instantiated), and knowable through ordinary observation/science — not
pure intellect like Plato thought.
Key vocabulary introduced:
• Realism about universals = the view that universals actually exist.
• Platonism about universals = the more specific view that universals exist and have
Plato's special features (transcendent, perfect, knowable only by intellect).
• Nominalism = the opposite view — denies universals (or abstract entities generally) exist
at all.
• Conceptualism = a middle position: universals exist, but only because our minds create
them by mentally abstracting features from objects — so they depend on minds.
• Important distinction: realists (of any kind, including Platonists) think these entities are
mind-independent — they'd exist even if no humans ever existed to think about them.
Even Plato, despite saying we need our intellect to know the Forms, still thought the
Forms themselves would exist with or without us.
The One Over Many argument begins: This is described as the classic, historically recurring
argument for believing universals are real. The book starts with a specific version from Quine's
essay "On What There Is," where Quine describes an argument he attributes to a character called
"McX":
1. There are red houses, red roses, and red sunsets. (Premise — several different things
share something.)
Therefore... (conclusion continues onto next page)
Page 9 (page 68)
This completes McX's argument and unpacks it.
Full argument:
1. There are red houses, red roses, and red sunsets.
2. Therefore: these things have something in common — the universal "redness."
What makes this a "One Over Many": You start with a premise about many objects that
share a similarity (the "many"), and you conclude there must be a single universal (the "one")
that explains why they're similar. Basically: if several different things are all red, there must be a
single thing — "redness" — that they all share.
This argument is ancient — traced back to Plato's Republic (~360 BC): The book quotes a
passage where Plato argues that whenever we use one word ("couch," "table") for many different
objects, we're committed to a single Form or idea behind that word. So many couches imply one
Form of "couch," many tables imply one Form of "table."
A worry raised: If you accept this reasoning for every common word (not just "couch" and
"table," but also "red," "round," etc.), you end up generating an enormous number of universals.
Someone who likes Ockham's Razor (the principle: don't multiply entities beyond necessity —
keep your theory as simple as possible) would be uncomfortable with that. This leads some
philosophers to think the argument should be limited to fewer cases.
The general form of the argument is given:
• Premise: There's some object x that is F, and some different object y that is also F.
• Conclusion: There exists a universal "F-ness" that both x and y share (instantiate).
(Here "F" is just a placeholder for any property, like "red" or "round.") McX's original version
used three examples (houses, roses, sunsets), but really you only need two distinct things sharing
a property for the argument's basic structure to apply.
Page 10 (page 69)
This page shows several different versions of the "One Over Many" argument to demonstrate
how far you could stretch it if you allowed it for any shared term.
Examples given:
• Red: There are red houses and red roses → there's a universal "redness" they share.
• Chair: There are two chairs in this room → there's a universal "chairness" they share.
• "In This Room": There are people and coffee mugs in this room → there's a universal
"in-this-room-ness" they share.
• Mass of one kilogram: Flowerpots and handweights both weigh one kilogram → there's
a universal "mass of one kilogram" they share.
The point of listing these: If you're as generous as Plato seemed to be — accepting a universal
for every term used to describe a group of similar things — you end up believing in a huge
number of universals, even silly-sounding ones like "in-this-room-ness." This generous approach
is called an abundant theory of universals: a version of realism that posits a large number of
distinct universals — in the most extreme version, one for literally every term that applies to
more than one thing.
Page 11(page 70)
This page introduces the opposite, more restrictive approach.
Sparse theory of universals: Instead of allowing a universal for every shared term, some
philosophers (like Armstrong) argue we should only accept a universal when the objects are
genuinely similar in some deep sense — not just similar because we happen to use the same
word for them. A sparse theory posits a small number of universals.
Armstrong's extreme version: Universals should only correspond to things recognized by our
best fundamental physics — not by everyday language. Under this view, out of all the
examples on the previous page, only "Mass of One Kilogram" would count as a legitimate
universal, because mass is a real category in physics. "Chairness" or "in-this-room-ness"
wouldn't count — those are just convenient labels we use, not genuine deep similarities in nature.
A middle ground exists too: Some philosophers take a moderate position — more generous
than strict physics-only sparse theory, but not as generous as "any shared word counts."
• Example: they might accept that "redness" is a genuine similarity, because all red objects
share something real — they all look the same way to normal human eyes.
• They might also accept "chairness" as genuine, since all chairs share the real functional
feature of letting people sit on them.
• But they'd still reject "in-this-room-ness" as a genuine similarity — because completely
unrelated objects (a person, a mug, anything else) can just happen to be in the same room
by coincidence, with nothing deeper connecting them. So that phrase doesn't track a real
natural similarity, and this version of the argument fails.
So the overall thread across these four pages: Plato put universals in a separate perfect realm;
Aristotle put them inside ordinary objects. The One Over Many argument is the classic
reasoning used to argue universals exist at all — but philosophers disagree sharply about how
many universals that argument should commit us to, ranging from "one for every shared word"
(abundant) to "only what physics tells us is real" (sparse), with moderate positions in between.