QuestionsCan a machine think?Reasoning is nothing but reckoning
Can a machine think?Thinking is computing

Reasoning is nothing but reckoning

Reason as reckoning (Hobbes)

Thomas Hobbes 1588 · 1679Without a state, life would be solitary, brutish and short.
See profile
What it claims

Reasoning is nothing but calculating on signs, and a device that applies those rules correctly reasons, whether it is made of flesh or of gears.

Why

Reasoning is adding and subtracting signs chosen by convention; an operation that is only that can be done by a device.

In short

"Reasoning is nothing but reckoning" is an argument by Thomas Hobbes (Leviathan, I, 5 (1651); De Corpore, I, 1, 2) on the question "Can a machine think?" (answer: thinking is computing). Its claim: Reasoning is nothing but calculating on signs, and a device that applies those rules correctly reasons, whether it is made of flesh or of gears. According to AskPhilosophy's logic check (Z3 solver), the argument is valid only if one grants a premise the author leaves unstated (P3). AskPhilosophy presents 3 objections, 2 of them with a documented reply, and the original text behind each premise, checked word for word.

Leviathan, I, 5 (1651); De Corpore, I, 1, 2
Valid with an implicit premise

In 1651, Hobbes claims that reasoning is adding and subtracting, as we do with numbers, but with words. If that is true, reasoning thought is a calculation on signs. Hobbes does not himself speak of a machine that reasons, but he also writes that automata have an 'artificial life'. This is why he is often seen as the ancestor of the idea that a well-made machine could think.

P1To reason is to add parts to get a whole, or to subtract one sum from another to find a remainder. When done with words, it means following the consequences of their names.
P2The names we use are general signs, agreed by convention to mark our thoughts and make them known to others. The calculation therefore works on these signs.
P3 · implicit premiseAn operation that consists only in adding and subtracting signs according to rules can be carried out by any system that applies those rules, whatever it is made of.
So reasoning is nothing but calculating on signs, and a device that applies those rules correctly reasons, whether it is made of flesh or of gears.

click a premise to reject it

If you accept every premise, logic forces you to accept the conclusion. So the debate is about the premises.
Checked by a logic solver: the premises cannot all be true while the conclusion is false.
Without the implicit premise: Without P3, the computer finds a case where reasoning is indeed a calculation on signs, but only a living being can perform it: Hobbes's definition stays true, the conclusion about machines does not.
Valid with P3. P1 and P2 say what reasoning is (a calculation on signs). P3 says this calculation does not depend on what carries it out. Hobbes does not write it in this form; he only writes that spring-driven automata may have an 'artificial life'. Without P3, the conclusion about devices does not follow: reasoning could be a calculation and still require a living brain.
Analogy. A division is done the same way by hand, on a calculator or in the head: what matters is the rules, not the tool. Hobbes says that reasoning, too, is applying rules to signs.

Objections

3

Each objection is a conversation: click to see who replies and where the debate stands.

René Descartes 1637 · attacks premise 3

One can conceive a machine that utters words, but not that it arranges them variously to reply to the sense of everything said in its presence, as even the dullest men can do. And reason is a universal instrument, whereas a machine's organs need a particular arrangement for each action: a machine cannot have enough arrangements to act as reason does in every circumstance.

a defender replies
see the argument
We have not yet found a direct reply by Hobbes in the texts.
Alan Turing defender of the argument · 1950
Digital computers can mimic any discrete-state machine: it is unnecessary to design a new machine for each task. One digital computer, suitably programmed for each case, can do them all.
Noam Chomsky 1966
Chomsky takes up Descartes: the normal use of language is creative, unbounded and fitted to the situation without being caused by it. A universal machine can compute anything, but that does not explain this use.
The debate goes on (our summary) Descartes wrote before the Leviathan and Hobbes did not answer this passage. The debate went on with modern computers.
Where the debate stands (our summary)Our summary: the debate is about what a program can cover. Defenders say a universal machine needs no organ per situation; Descartes's heirs reply that what matters is the creative and appropriate use of language, which is not just applying rules fixed in advance. The core disagreement is unresolved.
Leibniz 1714 · attacks the conclusion

Even if reasoning were a calculation made by parts, perception is not explained by parts: walking into an enlarged machine as into a mill, we would see only parts pushing one another, never anything explaining a perception.

no reply
see the argument
No reply possible: objection raised in 1714, after the author's death (1679).
Where the debate stands (our summary)Our summary: Leibniz speaks of perception, not of calculation. Even if he is right, one can hold that reasoning is a calculation without calculation sufficing to perceive or understand. The debate continues in the linked argument: an open point.
John Searle 1980 · attacks the conclusion

A man locked in a room who follows a program to manipulate Chinese signs can give good answers without understanding a word of Chinese. Following rules on signs is therefore not enough to understand, or to think.

a defender replies
see the argument
No reply possible: objection raised in 1980, after the author's death (1679).
Daniel Dennett defender of the argument · 1987
Dennett argues that the experiment works because we imagine a trivial program. A program able to hold a real conversation would be immensely complex, and our intuition 'it understands nothing' becomes unreliable.
Where the debate stands (our summary)Our summary: Searle holds that manipulating signs does not by itself give meaning; his critics reply that the experiment makes us underestimate what a real conversational program would be. Core disagreement about what understanding and meaning are.