OgIC LOGIC witn numbers LOgIC in tibi.

ltUOe i ests
Ilock Logic and Truth Tables Algorith
klgebra Logic Equations Karnaugh Maps and Logic
tilts Introduction to Logic Types of Logic Deduc
iv is rams a Dia

uction
)eductive Logic Venn Diagrams Simple Logic Diar,
Inductive Logic - Logic with Numbers Logic
.itude Tests Block Logic and Truth Tables Algorithm
1, Boolean Algebra Logic Equations Karnaugh Mar
ind Logic
c De
)iagrams
n Aptitu e ables
klgorithms ions Ka
laugh Maps to Logic
hypes of Log n Diagrams
5irnple Logic Diagrams Inductive Logic Logic wit
lumbers Logic in Aptitude Tests Block Logic and Trut
rabies Algorithms Boolean Algebra Logic Equa
:ions Karnaugh Maps and Logic Circuits Introductio
'1 Logic Ty ot Logic Ven
w itivAl
agrams Si imuctive Logic
gic with Numbers Logic in Aptitude Tests Block Logi
id Truth Tables Algorithms Boolean Algebra Logi
luations Karnaugh Maps and Logic Circuits Introduc
)n to Logic Types of Logic Deductive Logic Ven
LOGIC
EASY
WARRIM

TAB BOOKS Inc.

Blue Ridge Summit, PA
 Published in the U.S.A. in 1985 by TAB BOOKS Inc.
Printed in the United States of America
First published 1984 by Lutterworth Press.
Reprinted by permission of Lutterworth Press. Contents

Introduction to Logic page

FIRST EDITION 5
Chapter
THIRD PRINTING 1. Some Types of Logic
9
Copyright © 1984 by TAB BOOKS Inc. 2. Which Type of Logic to Use?
19
Printed in the United States of America 3. Deductive Logic
25
Reproduction or publication of the content in any manner, without express 4. Venn Diagrams
permission of the publisher, is prohibited. No liability is assumed with respect to
the use of the information herein. 35
5. Simple Logic Diagrams
Library of Congress Cataloging in Publication Data 43
6. Inductive Logic
Warring, R.H. (Ronald Horace), 1920 - 49
Logic made easy. 7. Simple Arithmetic or
Logic with Numbers 56
Includes index. 8. Logic in Aptitude Tests
1. Logic, Symbolic and mathematical. I. Title. 65
QA9.W375 1985 511.3 85-12578
ISBN 0-8306-1853-8 (pbk.)
9. Introduction to Block Logic and Truth
Tables 71
TAB BOOKS Inc. offers software for sale. For information and a catalog, please 10. Algorithms
contact TAB Software Department, Blue Ridge Summit, PA 17294-0850. 81
11. Introduction to Boolean Algebra
Questions regarding the content of this book 85
should be addressed to: 12. More Logic Equations
95
Reader Inquiry Branch 13. Karnaugh Maps and Logic circuits
TAB BOOKS Inc. 103
Blue Ridge Summit, PA 17294-0214 Index
110
 Introduction
`Contrariwise', continued Tweedledee, ' If it was so, it might be; and if it
were so, it would be; but as it isn't, it ain't. That's
logic.'
from Lewis Carroll's Through the Looking Glass

That, as a concise explanation of logic, is quite clever. But Lewis

Carroll was not only an outstanding writer; he was also a mathemati-
cian and an active practitioner of logic, or logician. The
dictionary
defines logic much more broadly as the science of reasoning,
proof,
thinking or inference - which is not too logical when you come to look
into it. It merely hints at the various possible forms of
logic, which
really start with commonsense. That is the sort of logic most people
can understand, but it is not true logic. True logic follows rules which
must not be broken to justify logical conclusions or answers.
The first thing to observe when looking at the
subject of logic in
depth is that there are many different types of logic - so there is
for everyone to be able to find a particular hopt
type of logic which he or
she can readily understand and apply to reach
logical conclusions or
answers. This type of logic will then become a useful, everyday tool -
much more effective than mere commonsense.
This book has been planned with that in mind. It has been divided
into chapters which deal with different types of logic. Some will be
easy to understand; others possibly quite obscure at first. In the latter
case, skip those for a start. Read only those which immediately 'make
sense'. You can then start to apply that
type of logic right away. Go
back to the others later on, when you will discover that there is an
inter -relationship between all the different
types of true logic. Logic is
not just a subject for philosophers or mathematicians - it is
everyone can use and benefit from. something
Logic today, in fact, is a 'popular' subject. There are various
nals and publications, for instance, devoted jour-
solely to presenting
puzzles and problems for solution by 'logic' as
exercise or even hobby -interest -just as crossword
a stimulating mental
puzzlesattract their
6 Logic Made Easy Introduction to Logic
7
addicts. The same sort of 'logic' problems have also been set for years formal logic were subsequently developed by other
groups of philoso-
in aptitude tests. Largely, however, this is a back-to-back form of phers, of which probably Stoicism (Stoic logic) is the most
logic. The originator of the problem starts with the answer and derives outstanding
as the source of sentential calculus. Except for the dedicated student
the facts or clues to fit it, which are then presented to the reader to of philosophy, then there is little else of note until the middle of the
solve and produce the same answer. sixteenth century when the first of the 'mathematical
Logic, too, has become a powerful modern design tool, particularly logicians' began
to appear, culminating in the development of Boolean algebra almost
for the complex circuits used in microprocessors and computers; the simultaneously by George Boole (1815-64) and Augustus De Morgan
more complicated 'engineering' control circuits, etc; and, of course, (1806-71). Mathematical logic, in its modern form, is finally accre-
computer -aided design. All the modern forms of practical (mathe- dited to Gottlob Frege (1848-1925) with his consistent and
matical) logic deal in binary relationships, a tailor made language for complete
development of the sentential calculus; subsequently further de-
computers and the ever-expanding field of design by logic rather than veloped by Bertrand Russel and Alfred North Whitehead in Princi pia
by combersome equations and tedious mathematical solutions. At the Mathematica.
same time binary language, as expressed in Truth Tables and Boolean All this is concerned, basically, with the logic used by
algebra, is an extremely simple, useful tool for application to problems philosophers
(and of necessity has omitted many other important names in the
in 'ordinary' logic, where it can often be a much more compact (and development of philosophical logic). Simplifying it to a degree, Aris-
exact) alternative to deductive logic. The basic mathematical princi- totle remains the inventor and inspiration behind deductive logic, as
ples involved date back for more than a century. Before they had a it is still used today. Mathematical
logic based around sentential
definite practical application they remained of interest only to the calculus, was 'finalized', as it were, only at the beginning of the
`pure' logician concerned with academic studies. Bertrand Russell -a present
century and because of its complexity demanded a new artificial
leading philosopher/mathematician - even gave up his studies of language to support it. It thus remains largely the tool of philosophers.
mathematical logic because the solutions obtained were 'too exact'. Indeed, as far as philosophers are concerned, the
On the other hand, formal logic has been the delight of philosophers algebraic
approach to logic using binary relationships remained as a weaker
for two thousand years, with philosophers greatly outnumbering tool, although as noted previously it has now become an
mathematicians throughout. This has produced the logicians whose design tool. For that reason, if you go to your local library important
to find
proper business, as defined by a leading (American) authority is 'the books on logic, virtually all will be catalogued and displayed under
investigation and formulation of general principles concerning what Philosophy. Treatment of practical modern logic will only be found in
follows from what; and whether particular examples of his own reason- parts of books on the Electrical and Mathematical shelves.
ing are valid or not is irrelevant ... Correct reasoning, however It is to be hoped, therefore, that this present book will fill a real
praiseworthy, does not itself contribute to logic.' The first sentence of need for introductory coverage of all types of logic in a single volume;
this virtually repeats Tweedledee's definition of logic. The remainder and above all make these types of logic easy to understand and apply.
contains two statements which may seem surprising. 'Correct' logic
does not have to be valid (i.e. the reasoning behind it does not
necessarily have to be true!). Also if the reasoning is correct, it is not
necessarily logic! You will find these two apparent contradictions of
the meaning of logic explained in the chapter on Deductive Logic.
Historically, the origins of formal logic date back to the Greek
philosopher Aristotle (384-322 BC), who developed the basic theory
of syllogism, which has been the core of deductive logic ever since as
first developed by the Peripatetic school. Alternative approaches to
 CHAPTER 1

Some Types of Logic

Commonsense Logic
Commonsense logic is deriving conclusions from
personal experi-
ence and/or knowledge. A conclusion that something makes sense, so
it is right; or something does not make sense, so it is
wrong. Common-
sense does not necessarily produce
correct answers, however; and is
not necessarily 'logical' at all - especially when compared with deduc-
tive or mathematical logic, for example; or even established
facts.
Commonsense 'logic', for example, would maintain that
brick,
stone, metals are all hard and solid substances. Science establishes that
the atomic structure of any solid substance is almost entirely
empty
space. Commonsense finds it difficult, or impossible, to accept such a
fact. It is not understandable, so it is not real. Even less 'realistic' is
time -dilation in space travel; or the quantum theory which holds
that
everything can be reduced to and analysed in terms of wave forms.
Yet commonsense is the logic most individuals use for
solving ordi-
nary, everyday problems. Applied to pure logic problems it can even
provide instant answers which are known to be correct,
rules of logic and avoiding any necessity of ignoring any
positive proof that the
conclusion is correct.
A
person familiar with philosophical logic and deduction would
immediately identify the following as a syllogism with an invalid
conclusion:
All dogs are animals
All cats are animals
Therefore all cats are dogs.
He could further go on to prove that this is an invalid
drawing a Venn diagram, or by showing that it breaks syllogism by
one or more of
the rules of deductive logic. Equally, he would
to argue this proof at some length. probably be prepared
Commonsense logic has not heard of syllogisms or valid or invalid
arguments. It simply affirms, without argument, that cats are not dogs.
 10 Logic Made Easy Some Types of Logic 11
To do this, however, it has to know there is a difference between cats Sherlock Holmes's Logic
and dogs. Without any knowledge at all of the French language, for Almost everyone must be familiar with the infallible reasoning of
example, commonsense logic could also maintain that all cats are Sherlock Holmes - his ability to put together a complete
chats is also wrong (because the spelling is wrong). the most meagre of clues whilst the amiable Watson picture from
was simply
This, incidentally, leads to an interesting digression. confused by the situation. Deductive logic at its best - or is it?
All cats are animals In fact it is not, although it is based on the principle of deductive
All chats are animals logic. It is fiction, written 'backwards' from the answer. Clever, ima-
Therefore: (i) all chats are cats ginative writing where the answer (conclusion) is first established, the
(ii) all cats are chats.
clues (premises) then extracted from the answer and then whittled
down to the absolute minimum to be ultimately acceptable for
True or false? (i) is true and (ii) is false. But this is a question of justi-
fying the answer. The possibility of alternative answers is not con-
knowledge, not logic. (In the case of (ii) all cats are not male (French) sidered for it does not fit in with the story, or the character of Sherlock
cats.) Holmes.
Here is a classic problem in commonsense logic. A bear walked one Let's face a simple situation which Sherlock Holmes'
mile due South, then turned to the left and walked one mile due East. type logic
would answer immediately with ease.
Then it turned to the left again and walked one mile due North and One Spring day you come across two round stones and a carrot
arrived back at its starting point.... What was the colour of the bear? lying close together on a grass verge. What do you deduce from that?
Now apparently the bear walked three legs of a square, like this The Sherlock Holmes type of mind would immediately answer: 'A
t1.But since it ended up where it started from its actual path boy built a snowman there in the second week of February.' (Sherlock
Holmes himself would probably have gone on to describe the
must have been a triangle A The only two places in the world
where this could happen is if the starting point is either the North Pole
boy in
more detail; where he went on holiday last (by the type of stones); and
even the snowman itself).
or the South Pole. The South Pole is ruled out since it is impossible to Now possibly this answer is right. Why a snowman? Because the
travel South from it. So the bear was at the North Pole, i.e. it was a stones were used for eyes, and the carrot for a nose. Why a
polar bear. So the bear was white. boy?
Because a girl would have taken more trouble and moulded the nose
Note again that this problem is solved by knowledge, not logic as such in snow. Why the second week in February? Beacuse that was when
- although it needs a logical type of mind to apply that knowledge to there was the last fall of snow that laid heavily.
a particular problem. So commonsense logic can perhaps best be None of these conclusions is supported by fact, so it is difficult to
described as logical reasoning based on the available facts, and drawing justify them on any logical basis. Indeed there are
on additional knowledge or experience to arrive at an answer. It could many other possible
answers, for example:
almost - but not quite - be called inductive logic, which is a recognized
category. It is certainly not true deductive logic, which does not permit (i) The two stones just happened to be there,
argument outside the facts available. fell out of someone's
anyway. The carrot
shopping bag at that particular spot.
In this particular example, too, logical reasoning can also be proved (ii) Digging up his vegetable plot, a gardener turned up two stones
by geographical fact - except for one thing. There is the remote and an old carrot. He threw them all
over the hedge, where they
possibility that a brown bear could have been taken to the North Pole landed on the grass verge close together. Neither
very interesting, but
by aircraft, say, as an environmental experiment. both equally as plausible as the 'snowman'
theory, or even more so.
(iii) One night a burglar approached the house the other side of
the
grass verge carrying two stones he could use to break a window
 12 Logic Made Easy Some Types of Logic 13
and force an entry. He thought the house was empty, but it was not; or come to a conclusion, can at first appear obscure because it is
so he turned, ran and jumped over the hedge on to the grass verge, wrapped up in a lot of ordinary language. The 'facts' of the
dropping the stones where he landed. The woman who was in the have to be extracted and put down in their simplest form, question
house and saw him was in the kitchen. She picked up the nearest eliminating
any ambiguity of duplication (synonymous description).
object, which happened to be a carrot and threw it at the burglar.
The carrot landed in the same spot as the stones. Reductio ad absurdum
That answer has 'written a story' around two simple facts. See how Reductio ad absurdum is a particular type of logic favoured
by philo-
many other different stories, or answers, you can think of. sophers, as well as being applicable to other types of logic. It
means,
quite simply, showing that an assumption is false by deriving a further
Ordinary Language conclusion from it which is absurd (i.e. cannot be true). In fact, it
can
Arguments in ordinary language are often difficult to prove, even be argued that reductio ad absurdum itself is absurd as
applied in deduc-
though they may lead to correct conclusions. This is the basic weakness tive logic since it argues at length about things which are not
true.
of ordinary language for dealing with problems in logic. Where proof Debating the question rather than seeking the answer. In mathemati-
of the validity or otherwise of the conclusion is necessary, the available cal logic, however, reductio ad absurdum can be quite precise. It can be
information may first need changing into what is known as a used to prove that something is impossible, i.e. that
a particular answer
standard -form categorical syllogism before it can be analysed fully is not a correct solution or method of solution; or that there
is no
under the rules of deductive logic. The term 'standard -form categorical answer to that problem.
syllogism' is extremely off-putting when first met, but basically means Suppose the problem is to complete a series of numbers using each
reducing the component propositions into three separate terms. The of the ten digits 0,1,2,3,4,5,6,7,8 and 9 once and
once only so that the
validity (or otherwise) of the argument is then simple to establish (see sum of such numbers composed is 100. Can it be done? The reductio
ad
chapter on Deductive Logic). absurdum approach will show that it cannot.
As an example, consider the following forms: Using the digits as' they stand, added together
No wealthy person:, are vagrants 0+1+2+3+4+5+6+7+8+9=45
All barristers are rich people
This is well short of the total required, so some of the numbers will
Therefore, no high court advocates are tramps have to be double figures instead of single figures to make
up the
These contain six terms, but in fact this is because of synonymous additional amount in tens. Call this sum T for tens. The sum of the
descriptions. Thus 'wealthy people' and 'rich people' are the same; so remaining figures is then 45 - T, so we get the following
are 'barristers' and 'high court advocates'; and so are 'vagrants and equation
expressing the requirement
tramps'. Thus, still using ordinary language, here is the same thing in 10 x T + (45 - T) = 100
standard -form containing just three terms.
No rich people are vagrants Solving this gives T = 55/9 = 6.1111....
This is obviously absurd. You cannot have 6.111...
All barristers are rich people
tens. T must be a whole number. So, by reductio addigits denoting
Therefore, no barristers are vagrants. absurdum, the
problem is not solvable.
That is now a standard -form categorical syllogism, which is easy to By all means apply reductio ad absurdum to deductive
but be prepared for a lengthy debate if logic as well -
pr,,ve as valid argument.
In discussing it with someone else.
At the same time this demonstrates a basic rule applicable to all philosophical logic there can be different opinions as to what is
types of logic. The information necessary to be able to solve the problem, absurd or not!
 14 Logic Made Easy
Some Types of Logic 15
Indirect Proof
which can be grouped like this
Indirect proof is closely related to reductio ad absurdum, but it works
in the opposite sense. It 'proves' a deduction or assertion by establish- 2X= (2-1) + (I -i) -i+ -I) -
ing that the opposite assumption is false. A handy trick for politicians, =1 and so on
this, to prove their policies right by expounding at length on how the which is back to the original series, thus
Opposition is wrong. 2X = X
Even more so than reductio ad absurdum, indirect proof is open to
or 2 = 1
objections when applied to philosophical logic, but again in mathe-
matical logic it can be quite positive. Proving that this mathematical reasoning is wrong can be quite a
Here is a simple example. If we add all the numbers possible problem!
together (i.e. 1 + 2 + 3 + 4+ 5 + etc., etc.) we conclude logically that No - the fallacy of this proof is not in grouping alternate pairs of
there is no end to such a series and so the sum of all possible numbers fractions and simplifying each group to a single fraction. Since the
is an infinitely large quantity - normally called infinity, or designated series extends to an infinite number of fractions, every fraction we
co. Thus 1 +2+3+4+5+ etc., etc. = co `borrow' from farther on in the series does not reduce the number of
Suppose, now, we adopt the opposite assumption that there is an end fractions in the series - nor does it extend infinity beyond infinity.
to such a series. If that is so, a further 1 can be added to it to give:
(1+2+3+4+5+ etc.,etc.)+1= oo+1 Lateral Thinking
But there cannot be a quantity co + 1 because co (infinity) is already Lateral thinking is another form of mental discipline which has re-
ceived considerable recent exposure via Dr de Bono's books - and
an infinitely large number (there cannot be a larger number). So
oo + 1 is impossible. Thus this second (opposite) assumption is false. particularly his TV series on the subject. Loosely, it is the art of
Hence the original assumption must be true. thinking problems through in more than one direction at a time.
Another form of logical argument, in fact, but only marginally ac-
`Proving' that 2=1 ceptable as a form of logic.
There are quite a number of ways of 'proving' mathematically that Probably the basic principle involved is best illustrated by example,
2 =1. Indirect proof immediately asserts that any such 'proof' is false. where something is being designed. The designer develops the
original
Sometimes this can be shown directly, i.e. proved false, without having idea through his own knowledge and experience into
something final,
to look too far. The trick of introducing co + 1 in an equation is one which is then built or produced. To complete it, it utilizes 'standard'
common method of 'proving' 2 = 1. In some cases, though, it can be parts bought in from another company, variation in performance or
difficult to see where the 'proof' is wrong. quality of which may affect the performance of the final product.
Take the following series of numbers added together, for example, Unless the designer has indulged in some degree of 'lateral
to take into account such possible effects (which were outside thinking'
calling the sum X: his
immediate design problem), the design could prove unsatisfactory.
X=1 +i +i -i+ and so on indefinitely. Here is a very simple example of this. The call is for a bracket to be
Now multiply by 2 to give a second series designed to hold two parts at a V -angle when screwed or bolted
the bracket. up to
Accordingly the design office drew up a V-shaped
2X=2x1-2xi+2x1-2x1+2xi-2xi+ bracket, and somebody in the workshop makes it. Then they find that
which gives they cannot possibly get long enough screws or bolts into the holes
near the bottom of the V -bracket. So a little 'lateral thinking' at the
2X=2-1+i-i+i-i+ design stage would have taken into account that not only was the
16 Logic Made Easy Sbme Types of Logic 17

shape of the bracket important, but the fact that it had to be fitted with adapting the style of natural language has been used by logicians since
bolts. Too bad if they had produced hundreds of brackets before the earliest days, although the first successful attempts to produce a
someone found that they were not usable! complete artificial language on this basis is due to the German logician
Gottlob Frege and only appeared towards the end of the last century.
Heuristic Reasoning The idea has since been developed extensively, replacing words with
Heuristic is the name given to a certain form of logic (or possibly symbols, and developing theorems of logic.
more correctly philosophy) involving a study of the methods and rules The aim is very simple - to give each expression an exact meaning,
of discovery and invention. Put in everyday language it could be free from context (which can often confuse or produce ambiguity of
described as arriving at a plausible guess to the answer of a problem. meaning in a natural language); and manipulate such expressions in
Heuristic reasoning is not regarded as taut (i.e. exact), but somewhat a logical manner (determined by rules and theorems). In practice this
provisional. It may (that is, the answer) be correct or incorrect. becomes a vast subject on its own, far beyond the scope of this book,
Equally, it may be based on inductive logic, or analogy. In either case and so is only mentioned as such. After all, the title is Logic Made Easy,
it can be a useful mental exercise, but heuristic reasoning is not and translating a natural language into an entirely new artificial
acceptable as positive or absolute proof. language is not easy. Just as it is not easy to learn computer language,
although with the great demand for such knowledge there are now a
Analogy considerable number of simplified computer languages which are
Analogy can be defined as a branch of heuristic reasoning and (relatively) easy to learn. Millions of people now use computers. The
defined simply as 'a sort of similarity'. From such similarity it is greatest number of computers in use, indeed, are the home computer
possible to make an educated guess at possible unknowns. However types. The number of people seriously interested in a comprehensive
the application of analagous argument is much broader than simple formalized logic language probably barely runs into thousands. Hence
argument. Analogies can be vague, or clear cut. They can be applied this type of language remains largely the prerogative of academics.
to mathematics, philosophy or other forms of argument. In many For instance, you start by translating a simple premise:
cases analogous argument and deduction is the only answer to lack of some A are B
knowledge in tackling a problem. Probably the clearest example here
is the case of a mathematical problem where a solution is quite into something like:
straightforward (and exact) using calculus, but the person's know- (3x) (Px & Ox)
ledge of mathematics does not extend to calculus. Using analagous
reasoning he (or she) could well derive an acceptable answer, if not and a little later on may find you have produced an expression
necessarily exact. something like this!
That, in fact, is the weakness of analogy. It is never exact argument
(x) (y) (P ad Oy ) & (Sxa, v Sya, x)
and deduction - it is heuristic reasoning.
(3z) (3w) ( (Lxz & Lyw) & (Pz & Pw) & (Sza2w v Swa2z))*

Logic Languages Nevertheless, you will find that this book does make some consider-
One of the basic difficulties in developing an approach to logic is able use of artificial language using symbols, but in a very, very much
the limitation of ordinary language when it comes to presenting prem- simpler form.
ises and conclusion in a formalized manner. (Exactly the same problem
*
arises with computers. You cannot instruct or 'talk' to computers in Incidentally, the natural language translation of this is - Tor every pair of
ordinary language - it has to be a formalized language consistent with prime numbers differing by 2, there is a pair of greater prime numbers
the input/output capabilities of the computer). So formalizing or differing by 2'. There are much simpler ways of expressing the same thing!
18
Logic Made Easy

Sentential Calculus
For the same reasons that the academic formalized language of
logic is only briefly mentioned, sentential calculus is also glossed over. It CHAPTER 2
relates to the artificial language and, broadly speaking, is based on a
series of metatheorems or proofs of general principles characterizing the
language. Every sentence consistent with a metatheorem is then either Which Type of Logic to Use?
a tautology (i.e. completely unambiguous) or a generalization of a
tautology. Everyone is born with the inherent ability to do some things well
To whet your appetite - or put you completely off the subject - here and also to suffer a sort of 'mental blank' in dealing with some other
is the start of metatheorem 110 in sentential calculus: subjects. Someone who is good at languages, for example, is often
quite hopeless with mathematics, or mechanical subjects. Some people
reason best with words, others with 'pictures' or diagrams.
If it looks interesting you can gain some further encouragement from It is just the same with logic. Some people will find one type of logic
the fact that there are only 22 theorems in the 100 -series covering the easy to understand and apply, but find other types of logic impossible
sequential calculus! to comprehend. So the answer to this particular problem is to work
with the type of logic you understand best. At the same time, though,
it does not necessarily follow that every type of logic can be applied to
solving any problem in logic. Often the reverse is true. Solving a
particular problem may involve a particular type of logic being used.
Only mathematical logic, for instance, will give exact answers to purely
mathematical problems.
Let's see how alternative types of logic can be applied to solving the
following problem:
George Brown, Tom Green and Bill White were talking. 'Funny
thing', said Brown. 'I've just noticed we're wearing different
coloured hats the same as our names."Yes', said one of the others.
`But none of us is wearing a hat the same colour as our name. For
instance, I'm wearing a green hat.'

Solution by Logical Reasoning

First write down the known facts or premises.
(i) Brown, Green and White are each wearing a coloured hat.
(ii) The colours of the hats are brown, green and white.
(iii) Brown pointed out this fact.
(iv) The one wearing the green hat then pointed out that none of
the colours of the hats they were wearing was the same as their
names.
All these premises are true.
20 Logic Made Easy Which Type of Logic to Use?
21
Analysis or argument: Since none is wearing the same colour as their The rest of the answer then follows in an obvious manner. The main
name: point is that the mathematical solution is shorter, quicker, and posi-
tive.
(a) Brown is wearing either a green or white hat
(b) Green is wearing a brown or white hat Incidentally, a complete proof of this solution is given in the chapter
on Boolean Algebra.
(c) White is wearing either a brown or green hat
(d) Either Brown or White spoke last, and is wearing a green hat.
Solution by Logic Diagram
Conclusion: Since Brown had already spoken, it must have been White In this case the problem is set down in the following diagrammatic
who spoke last (consistent with (d)). That means it was White who form:
was wearing a green hat (consistent with (c)).
That leaves Brown wearing a white hat (consistent with (a) since the brown green white
green hat has already been allocated). That leaves Green wearing the hat hat hat
brown hat (consistent with (b) since the white hat has already been BROWN
allocated). GREEN
WHITE
So - Brown was wearing the white hat
Green was wearing the brown hat
White was wearing the green hat Since the hat colours are not the same as the names
we can fill in
part of the diagram thus, using 'X' to show 'not':
Now the same problem can be solved by several other different
types of logic. If the following alternative solutions are not clear, read
brown green white
the appropriate chapter describing that type of logic first. hat hat hat
BROWN X
GREEN X
Solution by Simplified Boolean Algebra WHITE
Here we can allocate capital letters for the names - B for Brown, G X
for Green and W for White; and lower case letters for the colours of
the hats -b for brown hat, g for green hat and w for white hat. To proceed further we need another diagram to analyse the other
available facts:
Let's concentrate on Brown, as the first name in the list. There is no
need to write out the argument in full to mull over; we can express the spoke spoke
relevant facts in simple equation form, noting that a over a letter first last
means 'not'. BROWN
B = b or g or w (Brown is wearing a brown, green or white hat) GREEN
B = b (Brown is not wearing a hat the same colour as his name) WHITE
B= g (Brown is not wearing a green hat since the last person who spoke
is White) Brown spoke first, so, that completes the first line with a ,/ and X
The 1; cancels the b and the g cancels out the g, so (,/ for 'yes' and X for `not'). The person who spoke last
a green hat. He cannot be Green, so he must be White. was wearing
The second
B=w (Brown is wearing the white hat) diagram can thus be filled in like this:
 22 Logic Made Easy Which Type of Logic to Use?
23
spoke spoke brown
first last
BROWN 'I X
green OR WHITE
GREEN
WHITE X N./ white NOT

(Green did not speak at all, so we could fill in his line with X and X, C
but this is not necessary for solving the problem.) Fig. 2.1c
We can now go back to the original diagram and enter the fact that
White has been identified as the last speaker, wearing the green hat: The person who spoke last was wearing a
green hat, so this bit of
information in block logic is:
brown green white
hat hat hat spoke last-s. green 'Input' -0. BROWN or WHITE
BROWN X X
GREEN But it cannot be Brown since he spoke first, which puts a NOT in the
X
green 'input' line of diagram A:
WHITE X / X
brown
That leaves only one possibility for Brown. He must be wearing a NOT H
white hat. The completed diagram also confirms that Green is wearing
the brown hat: green --I NOT H OR BROWN = white hat

brown white
green white
hat hat hat
BROWN X X Fig. 2.2
N/
GREEN / X X The rest of the solution then follows by completing the
WHITE X ./ X GREEN and WHITE (i.e. introducing a second NOT in diagrams for
the appropri-
ate 'input' lines).
Solution by Block Logic Incidentally, rendered in Truth Table form, as the complementary
Here the possibilities are written down in the following form: method of expressing block logic, the solution follows
lines as the solution exactly the same
by Logic Diagrams, using 1 and 0 instead of ,/
and X.
brown H NOT H brown

green OR - BROWN green H NOT .' GREEN

Solution by Arithmetical Logic
OR
Not applicable since no numerical values are involved and there is
white white
no numerical answer.

A B Solution by Inductive Logic, Analogy or Heuristic Reasoning

Not necessary since all the facts
Fig. 2.1a,b necessary to obtain the solution are
available.
24 Logic Made Easy

Solution by Reductio ad Absurdum or Indirect Proof

This could be used to find an answer by assuming an answer and
seeing if this contradicts the facts. For example, assume Brown is CHAPTER 3
wearing the green hat. From the facts available, Brown spoke first.
But the person who spoke last was wearing the green hat. Thus the
assumption is not correct, i.e. Brown cannot be wearing a green hat. Deductive Logic
Equally, from the facts, he is also not wearing a brown hat. Therefore
he must be wearing a white hat. The rest of the solution then follows. In using the description deductive logic, the term deductive refers to the
This solution has been arrived at surprisingly simply. Note, how- manner of reasoning or argument. It follows the same pattern as all
ever, it has used a mixture of logic. The final solution, starting with the approaches to logic in listing a number of premises (or statements in
conclusion that Brown must be wearing a white hat, is derived by simple terms), which are then considered or 'argued' over to arrive at
deductive logic. a logical solution to the problem or conclusion. Strange as it
may seem
at first, the 'logic' part is concerned only with the correctness
or validity
of arguments - not whether the premises themselves are true or false, or
even whether the conclusion is true or false. But if the
argument is
valid (note: the argument is never itself said to be true or false), then
the conclusion must be true if the premises are true. If the
argument itself
is logically incorrect or invalid, there is a possibility that the
premises
could be true and the conclusion false, or vice versa. Let's clarify this
with some simple examples.

Example IA
True premises and true conclusion
Premises: All animals are mortal (true)
All cats are animals (true)
Conclusion: All cats are mortal (true)
This is valid argument.

Example 1B
On the other hand we can have true premises leading to a true
conclusion, but with invalid argument, e.g.
Premises: All animals are mortal (true)
All cats are mortal (true)
Conclusion: All cats are animals (true)
This is invalid argument, even if the conclusion is true, because there
is nothing in the premises to establish the fact that cats are animals. In
`knowing' that they are, we are assuming another premise, not
speci-
fically given.
 26
Logic Made Easy Deductive Logic
27
The logical correctness or incorrectness depends solely on the rela-
Example 3
tion between premises given and conclusion. You cannot 'add in'
False premises and false conclusion
another premise, however convenient. On the other hand, logical Premises: All cats have wings
argument can ignore a false premise, as in the next example. (false)
All dogs are cats (false)
Example 2 Conclusion: All dogs have wings (false)
True and false premises and true conclusion This is still valid argument since the conclusion reached is
Premises: Al} cats are mortal (true) a logical
assessment of the facts presented in the premises, even if these are false.
All birds are cats (false)
Conclusion: All birds are mortal (true) Example 4
This is valid argument since its conclusion reached is true, even if one False premises are true conclusion
premise is false. Premises: All cats have wings (false)
At this point it should be pointed out that although the examples All birds are cats (false)
quoted are simple statements this does not mean that deductive logic Conclusion: All birds have wings (true)
is restricted to simple, straightforward premises. Quite the reverse.
Valid argument again!
The original subject may be quite complex and lengthy, the gist of At this point, deductive logic may appear to be anything but
which then has first to be extracted in the form of simpler premises. If
but the above established the ground rules - the simplest and logical,
easiest
dealing with more than one subject, it would also have to be broken
down in order to provide conclusions for each subject separately. to understand being: If the premises are true, a valid argument must give a
true conclusion.
Simple statements are thus the end product of a preliminary assess- This then raises a most important point, that the form of the
ment of the subject matter, as a preliminary to applying deductive argu-
ment must also be valid for the argument itself to be valid. To explain
logic. They express the gist of what is being analysed in the simplest this simply, let us take the premises of example IA:
possible terms, ignoring irrelevancies.
One thing needs to be made quite clear. Not every statement is All animals are mortal
necessarily a proposition or premise in logic. To qualify it must be All cats are animals
capable of being either true or false. This can help explain some of Now cross-link the two statements:
the paradoxes in 'classic' logic, like the logic puzzle (antinomy) of
Bertrand Russell's: All animals are mortal
`In a certain town there is a barber who must shave all those people
and only those people who do not shave themselves.' This leaves the All cats are animals
question of whether the barber shaves himself or not. If he does, he
breaks the rule and so he must not shave himself. If he does not shave This reads as 'all cats are animals (and) all animals are mortal'
himself, then he is again breaking the rule and so he must shave which is a valid form of argument for concluding 'all cats are mortal'.
himself. In baisic 'formula' terms, this becomes:
The answer to this is that the rule itself is wrong. It is both true and All A are B
false in the sense that it both can and cannot be obeyed. (That can be
arrived at by deduction.) There is another possible answer, perhaps an
even better one. Logically, the barber cannot exist. (That can be All C are A
arrived at by intuitive logic.)
(using A for `animals'; B for 'mortal' and C for `cats').
 28
Logic Made Easy Deductive Logic
29
Now look at Example 1B, using letters again, which
only cross -links Further valid forms are:
like this:
(i) If A, then B
All A are B
Not B
Conclusion: Not A
All C are B
Example: Premises: If it rains, then I will get wet
The only common feature is that both animals and cats are mortal. I have not got wet
It is an invalidform to make a conclusion that all C Conclusion: It is not raining
are A (all cats are
animals), even if the conclusion is correct. Substitute 'fish' instead of (ii) If A, then not -B
`cats', for example: B
Conclusion: Not -A
All mammals are mortal Example: Premises: If it is fine, then I will not get wet when
All fish are mortal
I go out for a walk
Conclusion: All fish are mammals. I have gone out for a walk and got wet
Conclusion: It is not fine
The conclusion in this case is patently not true, because the form
of (iii) If not -A, then not -B
the argument is invalid.
Not -A
This, in fact, is one way of finding out whether the form of the
Conclusion: Not -B
argument is valid or invalid - see how they crosslink, and if 'substitu-
tion' to give an equally true premise (or a counter Example: Premises: If it is not raining, then I will not get wet
-example, as it is It is not raining
called), results in a conclusion which is obviously false.
This also explains why a true conclusion can be derived Conclusion: I will not get wet
even when Invalid forms of argument are:
one or all of the premises are false - provided the form of the
is valid. argument
(iv) If A, then B
B

Conditional premises Conclusion: A

A conditional premise is in the form Example: Premises: If Bacon wrote Shakespeare, Bacon was
If A, then B a great writer
Bacon was a great writer
Which is then followed by a second premise A. The valid form Conclusion: Bacon wrote Shakespeare.
of the
argument is then cross -linked like this (v) If A, then B
Not -A
If A, then B Conclusion: Not -B
A Example: Premises: If Bacon wrote Shakespeare, Bacon was
a great writer
As a simple example: Bacon did not write Shakespeare
Premises: If it rains, I will get wet Conclusion: Bacon was not a great writer.
It rains The validity, or otherwis
Conclusion: I will get wet e, of premises and conclusion can also be
proved
mechanically, using Truth tables or Venn diagrams.
 30
Logic Made Easy Deductive Logic
31
Categorical Statements (i) The middle term must be distributed exactly once
`Categorical' means unconditional, so a categorical statement is (ii) No end term may be distributed only once
explicit, e.g. 'All A are B', or 'No C are D'. The former is an affirmative (iii) The number of negative premises must equal the number of
statement and the latter a negative statement, but both are explicit or
negative conclusions.
universal as far as logic is concerned. But they are not necessarily
expressing an unconditional truth. The statement 'all trespassers will
be prosecuted', for example is categorical, but it does not To apply, these, however, we have to understand clearly what
necessarily `distributed' means, and to do that we first have to be able to identify
mean that there will be trespassers - and the validity of being able to
the subject and predicate in each categorical statement. Distribution
prosecute them if there were is something that does not necessarily
hold true in law. then follows from whether the statement is universal or
particular,
Other categorical statements in logic can be even more ambiguous affirmative or negative (refer back to the start of the section on
or even contradictory if they are of the type: `Categorical Statements').
If the statement is universal and affirmative, then the subject is distri-
Some A are B; or in negative form, Some A are not B buted and the predicate undistributed. If the statement is
particular
These are known as particular statements, thus proving that a categorical and negative then the subject is undistributed and the
predicate distri-
statement can be a true premise can be difficult at times. One way of buted.
doing this is to deny or 'invert' the statement, like this If the statement is universal and negative, then both the subject and
predicate are distributed.
statement: A
If the statement is particular and negative then the subject is undistri-
re -write as Not -A
buted and the predicate distributed.
Now if Not -A is shown to be true, the original statement A is false. On Don't be confused at this stage. We can make everything much
the other hand, if Not -A is false, then A is true. In other words, the simpler (and more understandable -logic) by going back to our syllog-
statement itself is argued in valid logic. Equally, this is the basis of ism example:
debate, where conclusion is finally decided by vote rather than truth!
All cats are mammals
Syllogisms
All mammals are animals
Syllogisms are arguments comprising only categorical statements All cats are animals
and embracing two premises and one conclusion - the basic formula
for deductive logic, in fact. Thus our original example IA is a syllog-
ism, for instance, and also the following: and re -write in simple 'formula' form using S for
subject, P for predi-
cate, and subscripts d and u to indicate distributed and
Premises: All cats are mammals undistributed
terms, respectively. We also need to identify the middle term
All mammals are animals which we will designate M. separately,
Conclusion: All cats are animals.
Now each of these categorical statements contains two terms, one of All cats are mammals
which is a subject and one of which is a predicate. One term occurs in
becomes SdK, (universal affirmative)
each premise (`mammals' in the example) and is called a middle term. All mammals are animals
Each of the other two terms occurs once only in the two premises and becomes MdP,, (universal affirmative)
once in the conclusion (`cats' and 'animals' in the example) are known All cats are animals
as end terms. The rules of logic then state that for a syllogism to be valid: becomes S,Mu (universal affirmative)
 32
Logic Made Easy Deductive Logic
33
Extract the 'formula' on its own
Some diamonds are not valuable (particular
S,M Conclusion: Some diamonds are not gemstones negative)
MdPU (particular negative)
Sd Pu
Pd M
Su Mc,
(Now refer back to the definitions for the rules for a syllogism to be
valid.)
SP,
Rule I is satisfied because the middle term is distributed Check Rule III first. There is
exactly once one negative premise and one negative
(in the second line) conclusion, so this rule is satisfied.
Rule II is satisfied because the first end term (S) is distributed Now check the other rules.
twice; Rule I is satisfied.
and the second term P is not distributed in either of its
Thus no end term is distributed only once. appearances. Rule II is satisfied.
Rule III is satisfied because there are no negative The syllogism is therefore valid.
premises and no
negative conclusions. Deductive logic can, quite easily, introduce
instance, is a classic example. ambiguity. Here, for
The syllogism is therefore valid.
Now look at this syllogism:
Ambiguity
All men like football A teacher of law made a contract with one of his
(universal affirmative)
No women are men pupils that the
(universal negative) pupil would not have to pay for his lessons if he did not win his
Conclusion: No women like football first
(universal negative). case. The pupil completed the course of lessons, but did not take
In 'formula' form this becomes: cases. The teacher then sued for any
payment.
Md Pu The pupil analysed his position
logically, based on the following
Sd Md premises:
Sd Pd (i) Either I will win the case, or lose it.
Rule I is broken because the middle term is distributed twice. (ii) If I win my case, I will not have to
pay (the teacher will have
Rule II is broken because the end term P is distributed lost his suit for payment).
exactly once. (iii) If I lose my case, I will not have to
Rule III is satisfied since there is one negative pay since this is implicit in
premise and one
negative conclusion. my contract with my teacher.
The syllogism is thus invalid. (You could also deduce this from the fact Since all three premises are true, the logical conclusion
win or lose my case I will not have to is that whether I
that both premises are false, but that does not pay.
justify assuming that In a similar manner the teacher analysed his
the conclusion is also false.)
these premises: position, based on
Actually there is no need to proceed past the first rule in this
case;
any one rule broken means an invalid syllogism. In fact, if there are (i) Either I will win the case, or lose it.
negative premise(s) or a negative conclusion, checking conformity to (ii) If I win my case, the pupil will have to
Rule III is the first thing to do. If this rule is broken the pay me.
syllogism is (iii) If I lose the case, the pupil will have to
invalid and there is no need to proceed further. pay me under the terms
Perhaps one more of the contract because he will have won his first case.
example to help your familiarity with syllogisms?
Again, since all three premises
All gemstones are valuable (universal affirmative) are true, the logical conclusion is that
whether I win or lose my case I will not have to pay.
34 Logic Made Easy

Somewhere there is a false premise, leading to a false conclusion, in

either the pupil's or the teacher's arguments. Or is there one in both?
Or are both sets of premises and conclusions logically correct? It is a CHAPTER 4
case of spotting the contradiction (s) or logical falsehood (s) - and
finding the answer is not as simple as it seems. Perhaps, indeed, it is a
subject for further argument and debate, with no logical conclusion at
at all?
Venn Diagrams
Venn diagrams - named after the nineteenth-century English logician
and mathematician John Venn - are a pictorial method of represent-
ing, and analysing, deductive logic. They can be simple and effective
to use - or confusing and difficult to understand. It depends, basically,
on how you react to the principle involved.
In a Venn diagram each particular class (or what is also called a
`set') is represented by a fully enclosed shape. It does not matter what
the actual shape is; it can be a circle, a square, a rectangle, or an
irregular shape, provided it is a closed shape. For the purpose of
explanation, elliptic shapes will be used for a start.
Any one shape then represents a class, e.g. in the first diagram of
Fig. 4.1, this represents 'all A', or mathematically A= 1. If the shape

A=1 A *1
i.e. A = 0
or A
Fig. 4.1

is shaded, as in the second diagram, this excludes all A from being

represented within this shape. Thus it represents A 0 1, or A = 0, or A
(not A). Enclose the shape within another shape as in the third
diagram and all A is now represented by the smaller shape (and
contained within it), so that everything outside and included within
the larger shape is not A. Equally, if we shade the smaller shape to
exclude A from it (making it A), then all A is contained within the
remaining plain area contained within the larger shape, as in the
fourth diagram.
Fig. 4.2 extends this principle to two states, A and B, where the two
shapes overlap. Thus in the first diagram this implies that some A are
 36 Logic Made Easy Venn Diagrams 37
diagram (Ili) Without shading, neither A nor B is excluded from the overlapping
area, i.e. `some A is B (diagram (iii)); and `some B is A' (diagram
(vii)). Since this is the significant part of the diagram this can be marked
with an X.
diagram (i)
Diagram (iv) shows a further relationship, with 'X' marking the
diagram (ii)
significant part of the diagram - `some A is not B' (some A lies outside
Fig. 4.2
the B shape). Similarly, diagram (viii) displays `some B is not A'.
B and some B are A, since both A and B occur in the overlapping area. Blank parts of a Venn diagram mean nothing on their own until all
At the same time A cannot occur outside its shape so is excluded from the available information has been entered, either by shading (de-
the non -overlapping part of the B shape. Complete labelling is then as noting exclusion from that particular area), or by an `X' or `some'
shown in the second diagram. If these overlapping shapes were in- relationship, or both. Thus exclusion is expressed by a universal state-
cluded within a larger shape, a further area would be produced from ment, e.g. `all A are ...'; and an `X' designates a particular relation-
which both A and B are excluded, i.e. A$ within this area as in the third ship, e.g. 'some A are ...'.
diagram.
Now look at the further diagrams shown in Fig. 4.3. These extend Venn Diagrams and Syllogisms
the principle of representing statements by Venn diagrams further. It Venn diagrams using two overlapping shapes, as just discussed, are
is to be understood that the left hand shape designates A and the right basically a diagrammatic representation of two categorical proposi-
hand shape B. By the rule established, A is excluded from the A shape tions. If a third overlapping shape is added the resulting Venn diagram
diagram (i) diagram (ii) diagram (iii)
can be used to test syllogisms, and/or solve problems in deductive
diagram (iv)
logic. This time, to establish the overlapping areas neatly the shapes
are best drawn as circles.
As a simple example, take the premises:
(i) All animals are mortal
All A isB No A is B Some A is B Some A is not (ii) All cats are animals
(In the chapter on Deductive Logic we have seen that the conclusion in
this case is 'all cats are mortal'; and the
argument is valid.)
Treating this in Venn diagram form we draw three overlapping
Animals Mortal
All B is A No B is A Some B is A Some B is not
diagram (v) diagram (vi) diagram (vii) diagram (vii
Fig. 4.3
by the shaded area in diagram (i) - which is then A - and only occurs
in the overlapping area with shape B. Thus all the A is contained in
this overlapping area, which is also included in shape B. Thus this
diagram expresses the condition or premise 'all A is B'.
Diagram (ii) excludes both A and B from the overlapping area
(shaded). That means 'no A is B'; and equally, `no B is A' - shown diagram (i) diagram (ii) agrarr. ;

explicitly in diagram (vi). Fig. 4.4

38 Logic Made Easy Venn Diagrams 39

circles, one for animals, one for mortal and one for cats, as in the first middle term must be distributed exactly once, i.e. appear unshaded in
diagram of Fig. 4.4. Now, all cats being animals excludes cats from each of the other terms (circles). Conventionally, too, the middle term
that part of the cat circle which is outside the animals circle. Shade (M circle) is placed at the bottom of the diagram.
that area in as in diagram (ii). The premises used in the previous examples have been universal
Equally, all animals being mortal excludes animals from the animal premises, i.e. of the form all A are B. The case of a particular premise,
circle outside the mortal circle, so shade that area in, as in diagram i.e. some A are B needs a little more explanation. For this, take the
(iii). We are left with the conclusion that the only area now accom- following which consists of one universal premise and one particular
modating all cats is within the mortal circle - hence 'all cats are mortal'. premise:
Now try an example where the premises are true but the conclusion
(i) All artists are gifted peop.e
drawn (by deductive logic) is false:
(ii) Some artists are poor
Premises: (i) All dogs are mammals Conclusion: Some poor people are gifted people.
(ii) All cats are mammals
Conclusion: All cats are dogs.
The Venn diagram for this is developed as shown in Fig. 4.6. The
first (universal) premise excludes all artists from the artist circle out-
This time shading -in the diagram produces the result shown in Fig. side the gifted people circle, so this part of the artist circle is shaded.
4.5 (last diagram). The only area where all the cats are contained is Only some artists are poor, though, so this area of overlap is marked
within the mammal circle. Cats are excluded from the dogs circle by with an X. This 'some' area is common to artists, poor people and
the shading. Therefore, from the Venn diagram, no cats are dogs. gifted people. Hence some people c-.-1 be artists (as given by the
Testing the (deductive) conclusion in this way shows that the argu- premise); and equally some poor people can be gifted (as given by the
ment in this case is not valid. diagram).

Dogs Mammals

diagram (i) diagram (ii) diagram (iii) Fig. 4.6

Fig. 4.5
In other cases the allocation of the 'X' area for 'some' may not be
clear from the premises, which will indicate that the argument is
Two premises imply three specific terms, each one of which is repre-
invalid (although the conclusion reached by deductive logic may or
sented by a circle in a Venn diagram. The conclusion can be tested by,
may not be correct). For this futher example, consider:
or deduced from, the final diagram. Thus a Venn diagram is essen-
tially a diagram or premises or propositions embracing a subject (S) Premises: (i) all good mathematicians are good scholars
or major term; a predicate (P) or minor term; and a middle term (M). (ii) some athletes are good scholars
A similar rule applies as for deductive logic - for valid argument the Conclusion: Some athletes are good mathematicians.
 40
Logic Made Easy Venn Diagrams 41
Athletes Mathematicians A A A A

B 2 3 1=Ag AB
2=AB
B AB
B 4
3=AB
4=Ae B Ag AB

Alternative form of
Good Carroll diagram
scholars Fig. 4.8

diagram (I) diagram (II) diagram (111) for each (A and not -A, B and not -B). The diagram then plots in full
the various possible combinations - four in this case.
Fig. 4.7 Unlike a Venn diagram, this type of diagram can be extended to
accommodate quite a number of separate terms before it becomes
In the corresponding Venn diagram - Fig. 4.7 - mathematicians
are excluded from their circle except where it overlaps the good excessively complicated. It is, in fact, basically the same as a Karnaugh
map - see Chapter 13.
scholars circle by virtue of the universal premise (i). Some athletes are
good scholars (from the particular premise), but there are two over-
lapping areas where this could be marked with 'X', as in diagram (ii).
There is not enough information in the premise to indicate whether
the 'X' should be in area or area 2. Area 1 seems the logical
1

placement, for it is known that some athletes are good scholars. But
there is no information to confirm whether or not some mathematicians
are also good athletes.. In this case the 'X' can only be placed on the
line between the two (it could belong to either area), as in diagram
(iii). The Venn diagram, in fact, shows that this form of syllogism and
argument is invalid.
Venn diagrams can also be used for mathematical solutions,
although there are simpler and more effective diagrams, or alternative
methods which can be used in such cases. The particular limitation of
Venn diagrams is that they are not very flexible. The greater the
number of terms to be accommodated the more unwieldy (and
pos-
sibly confusing) the diagram becomes. This virtually limits their
application to a maximum of four separate terms or variables (i.e. four
overlapping shapes).

Lewis Carroll Diagrams

Lewis Carroll - an excellent mathematician as well as author -
developed an alternative form of logic diagram using squares. Fig. 4.8
shows a Carroll diagram plotted for two variables and two conditions
 CHAPTER 5

Simple Logic Diagrams

Many people find it easier to solve logic problems by diagrams rather
than deductive reasoning or mathematical solutions - 'pictures' speak
louder than words. Basically, however, a logic diagram is a Truth
Table, complementary to block logic and Boolean algebra. But for the
non -mathematically minded person such a diagram can be divorced
from these more obscure subjects by plotting the diagrams on a 'yes'
or 'no' basis, using ,./ for 'yes' and X for 'no' instead of the more
formal symbols 1 and 0, respectively. A large proportion of problems
designed specifically as 'logic puzzles' can be solved by using diagrams
of this type.
Here is a simple puzzle -problem which can be solved using a logic
diagram:
Smith, Jones and Thomas live in London, Brighton and York, not
necessarily in that order. All travel away from home to work.
Smith travels to London to work.
Thomas lives further South than Smith.
Jones also works in London.
Find out where each lives.
The basic logic diagram is drawn up like this:
London Brighton York
SMITH
JONES
THOMAS
Smith travels to London to work, so he does not live in London. Put
an X against Smith under London.
London Brighton York
SMITH X
JONES
THOMAS
44 Logic Made Easy Simple Logic Diagrams 45

Jones also works in London, so does not live in London. Put an X More Facts - More Diagrams
against Jones under London. The remaining space in this column is Sometimes the facts available need treating in separate diagrams to
then filled with a tick - the only remaining possibility, i.e. Thomas solve alternative or complementary possibilities. Here is a further
lives in London. example illustrating this. Problem: There are three people with the
names Smith, Jones and Robinson. Their Christian names are Arthur,
London Brighton York Mary and Jane, not necessarily in that order.
SMITH X
(i) Their ages are 17, 24 and 30.
JONES X
(ii) Miss Jones is 7 years older than Jane.
THOMAS .\,/ (iii) The person named Smith is 30 years old.
What are their full names and ages?
Now Thomas lives further South than Smith. Since Thomas lives in
London that means Smith must live in York. The York column can (Note: after the original statement, the facts have been separated for
now be completed. The Xs in this column merely confirm that Thomas ease of reference. In a set problem the facts may be incorporated in
does not live in York (he lives in London); and since Smith lives in one complete statement. It is then first necessary to separate them out
as individual facts.)
York, Jones cannot.
To solve this particular problem, draw up diagram A relating
London Brighton York surnames to Christian names. The only immediate clue is that Miss
SMITH X Jones will not have the Christian name Arthur, so an X can be marked
N./ in the Arthur column against Jones.
X X
JONES*
THOMAS / X
SMITH
Arthur Mary Jane

The only possibility left is that Jones lives in Brighton, so the JONES X (diagram A)
diagram is completed like this ROBINSON

London Brighton York

There seems to be more information available about ages, so con-
SMITH X X N/ struct two more diagrams on this basis:
JONES
THOMAS /
X N/
X
X
X
SMITH
17 24 30
Arthur
17 24 30

Thus: Smith lives in York JONES Mary

Jones lives in Brighton ROBINSON Jane
Thomas lives in London.
Note the principle of working with such diagrams. Once a \/ (yes) (diagram B) (diagram C)
has been established in any horizontal line or column, then the remain-
Now enter the main facts:
ing spaces in that particular line or column can be filled with Xs.
Similarly, if all the spaces but one in a line or column are filled with (a) Smith is 30 years old - enter in diagram B and complete
Xs, then the remaining space must be a / (yes). SMITH line.
 46 Logic Made Easy Simple Logic Diagrams 47
(b) Miss Jones is 7 years older than Jane. The only possibility is
that Miss Jones is 24 and Jane is 17. Arthur Mary Jane 17 24 30
SMITH
Complete JONES line in diagram B and the Jane line in diagram
C. JONES
ROBINSON
17 24 30 17 24 30 17
SMITH X X Arthur
24
JONES X ,./ X Mary 30
ROBINSON Jane X X
(diagram B) (diagram C) This would fill in, step-by-step, in the same manner, transferring
Diagram B can now be completed: information from the side and bottom diagrams to the main diagram
as it becomes available.
17 24 30
SMITH X X \/
JONES X ,/ X
ROBINSON ,/ X X
Diagram C can also be completed by filling in the top line with the
only possible alternative (Arthur is 30)
17 24 30
Arthur X X \/
Mary XX ,/
Jane X /
X
From these completed diagrams we can now complete diagram A

Arthur Mary Jane

SMITH /
V X X
JONES X ,/ X
ROBINSON X X \/
Answer to problem: Arthur Smith is 30
Mary Jones is 24
Jane Robinson is 17
Instead of separate diagrams the whole problem can be entered on
one diagram, appending the additional diagram(s) required to the
right and bottom of the main diagram. In this case the logic diagram
would be:
 CHAPTER 6

Inductive Logic
Inductive logic is similar to deductive logic in that conclusions are
based on premises, but with one very important difference. The con-
clusions extend beyond the area of fact provided by the premises.
Similar rules apply as to what is valid argument and a true conclusion,
but the conclusion is necessarily qualified. It is not necessarily absolutely
correct, particularly as it often has to be based on incomplete facts. It
may be a convenient generalization, or even a forecast by extrapolation.
The 'safest' type of inductive arguments providing conclusions (i.e.
those least likely to be in error) are those based on observed data, but
even these have numerous pitfalls, correctly called fallacies. Take as an
example a spring suspended vertically and carrying a pan into which
weights can be put. The downward deflection of the spring is measured
with weights of 10, 20, 30, 40, 50 grams, etc., added to the pan and
plotted as points of a graph - Fig. 6.1. These points all lie in a line, so
inductive argument would conclude that the relationship between
deflection and load is a straight line, and the graph drawn in accord-
ingly. It would also seem logical that if this relationship is true it will

10 20 30 40 50
Weight in grams

Fig. 6.1
 50
Logic Made Easy Inductive Logic 51
extend beyond the measured range, i.e. downwards to give zero de- This point is made more clearly in Fig. 6.3, which shows a simple
flection with no load; and upwards to forecast the deflection produced wave form. Say this represents an alternating current, the value of
by larger loads. which is read by a meter. Suppose measurements are made at intervals
In fact, this conclusion holds true, up to a point. It could be used to 1, 2, 3, etc., which happen to correspond in time exactly to the peaks
calibrate a spring and a pan as a simple weighing machine, but there of positive current. This would give points lying in a straight line
are other factors (not catered for in the original premise) which could apparently indicating a steady positive current, Fig. 6.3 diagram (A).
modify the truth of the conclusion - Fig. 6.2. There could be a physical Equally, if they were made with the same interval but starting at a
quarter of an interval later ( 1', 2', 3', etc.), all the measurement points
would indicate zero current Fig. 6.3 diagram (B). Started half an
interval later, they would indicate a steady negative current.

15
0
!Spring
\/
weak
gets

positive t
0
diagram (A)

negative Measured at points 1,2,3 etc

2 3
Weight Weight -0.
Fig. 6.2 A 1 IA 2)
diagram (e)
I
I
Measured at points 1' ,2' ,3' etc
limit to the amount of spring deflection before the pan hits something, I

for instance. Or if there were no physical limit to movement, the spring ORIGINAL SIGNAL
could be stretched beyond its normal behaviour pattern (i.e. exceeds
diagram (C)
the limit of proportionality). Both would deny the truth of extrapo-
lating the original straight line graph upwards. Equally, the spring
could be affected by 'fatigue' in repeated use, so that it no longer
returns to zero deflection with no load. However, its behaviour within
the originally observed range is unaffected. Thus extending the con-
clusion too far beyond the area of fact has nullified the truth of the
conclusion. Fig. 6.3
In fact, even within the area of fact established by measurement,
the conclusion could still be false. In this case it is not, but with some This is an error (or what would be called in inductive logic terms a
other system where measured data spaced at regular intervals show fallacy) due to unrepresentative sampling. Not enough sample measure-
what is apparently a linear relationship (i.e. all the points can be ments have been taken to give anything like a true picture, so the
joined by a straight line), the true relationship could well vary differ- inductively reasoned conclusion is false. Take measurements at shorter
ently between points, e.g. instead of a straight line between points the intervals and this will immediately show up. Taken at one quarter of
true representation could be a curve waving up and down. the
original time interval, for example, the measured points and the
52 Inductive Logic 53
Logic Made Easy

subsequently induced plot would zig-zag up and down - Fig. 6.3 To be more complete the generalization would need to take in this
diagram (C). Still not true, but a much closer representation. second cause (temperature) into the final statement.
It is lack of sufficient samples (data), or unrepresentative samples A working 'formula' for such a method of reasoning would be -
which is a common reason for false conclusions in inductive logic, cause 1 (oxygen) in the presence of cause 2 (a suitable temperature)
particularly when applied to statistics. Opinion polls are a classic produced combustion of paper and wood (effect).
The other thing to emerge is that causal connections are based on
example. They are based on answers (data) provided by a limited
number of people, yielding, say, 65 per cent in favour of something a study of observed 'happenings' - what happens (i.e. what is the
and 30 per cent against (the other 5 per cent 'don't know'). The effect) in a particular circumstance or condition (cause). Equally
pollsters conclude, by induction, that 65 per cent of the whole population important is how many times this occurs. The more times the greater
the likelihood of the (generalized) conclusion being true, or at least
are in favour, 30 per cent against. As a matter of logic, this is basically
nonsense, unless the sampling is fully representative (which in fact it substantially true. This can be expressed in the following form:
can never be). It can only be a statistical generalization. It can also be Instance 1 of happening A is related to cause B present
biased unintentionally by poor selection of samples; or deliberately by Instance 2 of happening A is related to cause B present
drawing on samples most likely to give the desired result. Instance 3 of happening A is related to cause B present
This, indeed, is the chief limitation of inductive logic - basing a
conclusion on insufficient evidence (data) and/or jumping to a conclu-
sion (which is human nature, but not logic). At least these are honest
The conclusion to be drawn from this is:
errors (errors of ignorance), whereas biasing towards a conclusion is
not. All instances of happening A are accompanied by cause B.
In fact it is very difficult, or even impossible, to prove fully that any Note: This conclusion is reached partly by inductive reasoning and
inductive conclusion is true because it extends the conclusion beyond partly by analogy, induction and analogy being closely related.
the facts available at the time. It can become a 'theory of ...' or a Causal connections summarized in this way can at best be valuable,
hypothesis; but these in turn may become subject to further proof or and at least be merely suggestive. They can also lead to conclusions
disproof in the light of new evidence, or a new hypothesis. This applies which are not factually correct, but despite this may be acceptable to
particularly in the fields of mathematics and science. quite a number of people. Here is a classic example, using the form
given above:
Cause and Effect
Inductive reasoning can be called a study of cause and effect - Arthur broke a mirror, cut his hand, which was bad luck.
logicians call this causal connections. It is based on a condition (or cause) Mary broke a mirror, lost her purse, which was bad luck.
which is necessary to yield a certain effect, leading to a conclusion or Tom broke a mirror, sprained his wrist, which was bad luck.
hypothesis relating the two. As such it is an inductive generalization which Conclusion drawn: breaking a mirror causes bad luck.
may or may not be completely true.
Taking a simple everyday example, it is well known that paper or Testing Causal Connections
wood will burn in air, which contains oxygen. But it will not burn in There are, however, methods of testing causal connections and
a sample of air from which the oxygen has been removed. The induc- resulting conclusions, known as Mill's 'canons' (or rules), after the
tive generalization which could be drawn is that oxygen (cause) English philosopher of that name.
produces combustion (effect). However, this is only part of the true
(i) Method of Agreement
picture. To burn in the presence of oxygen the paper or wood must
first be raised to a suitable temperature, e.g. by holding a match to it. In words: if two or more instances of a happening under investigation
 54 Inductive Logic 55
Logic Made Easy

have only one circumstance in common, the circumstance in which (v) Method of Concomitant Variation
alone all the instances agree is the cause (or effect) of the given happen- In simple language, this accommodates varying happenings desig-
ing. nated by + or - signs and is easiest to understand in symbol form:
Represented in symbol form, using capital letters for circumstances
and lower case letters for happenings, this becomes: ABC abc
A+BC a+bc
A B C D occur together with a b c d A -B C a ±b c
A E F G occur together with a e f g Therefore, A and a are causally connected.
Therefore A is the cause (or effect) of a.
Although looking less understandable at first sight, this is actually
(ii) Method of difference the most useful of the methods of testing causal connections since it
In words: if an instance of a happening occurring and an instance can provide quantitative analysis of inductive reasoning.
in which it does not occur have every circumstance in common but
one, that one occurring only in the former; the circumstance in which
alone the two instances differ is the effect (or cause) or an indispensable
part of the cause of the happening.
In symbol form:
A BCD occur together with a b c d
B C D occur together with b c d
Therefore, A is the cause (or effect), or an indispensable part of the
cause of a.

(iii) joint Agreement and Difference

This combines both the Method of Agreement and Method of
Difference and in symbol form becomes:
ABC abc ABC abc
ADE ade BC be
Therefore, A is the effect (or cause), or an indispensable part of the
cause of a.
(iv) Method of Residues
Rather cumbersome in words, this is: subtract from any happening
such part as is known by previous induction to be the effect of certain
antecedents, when the residue of the happenings is the effect of the
remaining antecedents.
In symbols this becomes much simpler:
ABC abc
B is known to be the cause of b
C is known to be the cause of c
Therefore, A is the cause of a.
 Simple Arithmetic or Logic with Numbers 57

We have three different statements, so we can write three different

CHAPTER 7
equations, using R for red, G for Green and B for blue; and x, y and
z for the number of red, green and blue balls respectively. (A mathe-
matician would probably set about it that way.)
Simple Arithmetic or Logic with Numbers
Equation (i): (x times R) + (y times G) + (z times B) =60
Many problems involving numbers or quantities, can be solved by or using a . sign for multiplication:
simple 'schoolboy' maths. The first rule is that to find two unknown x.R+y.G+z.B=60
quantities there must be two separate statements relating those quant- There are 4 times as many red balls as green balls, so
ities.
Equation (ii): x= 4y
Example: Arthur is twice as old as Bertie. If their combined ages is
There are 6 more blue balls than green balls, so
54 years, how old are they?
There are two statements, so the problem is solvable. Equation (iii): z=y + 6
Write down the two statements in simple equation form, using A for
we now have three equations to manipulate to solve for three
Arthur's age and B for Bertie's age:
unknowns. They are simple equations in this case and easy to solve by
(i) A = 2B substitution, but the problem has got a little confused by introducing
(ii) A +B= 54 x, y and z. So let's start again with a logical approach.
Equation (i) gives the equivalence between A and B directly, so use There are more blue balls than green balls (one fact)
fact to rewrite equation (ii): There are more red balls than blue balls (another fact)
A+B=(2B)+B=54 That means that the green balls must be the least in number
or 3B = 54 (conclusion).
Hence B = 18 years Now let's put this together in simple arithmetical form, forgetting
And since A = 2B about the x, y and z approach, and using R, B and G for the numbers
A =2 x 18 = 36 years of red, blue and green balls, respectively. Use the number of green
balls (G) as a base, because they are the least number
This is a very elementary example but it is surprising how often the
equivalence can be extracted from one equation (statement) and directly number of green balls =G
substituted in the other equation (statement) to obtain the answer to number of red balls =4 times as many green = 4G
one of the unknowns. The answer to the other unknown then follows number of blue balls = number of green balls + 6 = G +6
quite simply, using either of the equations. Now together G + 4G + (G + 6) = 60
The method is readily recognizable as solutions to miscellaneous or 6G =60 -6
equations - a subject that clicked or not at school. But what happens = 54
when more than two unknowns are involved? Does this simple mathe- hence G = 9
matical method work? Not necessarily, unless you apply a little logic,
It then immediately follows that
as the following will show:
There are a mixture of red, green and blue balls in a box. The R=4G=4 x 9=36
total number of balls is 60. There are four times as many red balls B=G+6=9+6= 15
as green balls; and 6 more blue balls than green balls. The point of this example is not to over -complicate the problem by
58 Logic Made Easy Simple Arithmetic or Logic with Numbers 59

writing out comprehensive mathematical equations. Approach the (vi) By the same reasoning, the next highest possible total is 45,
problem logically, looking for the simplest solution. The original which gives the following answers:
mathematical equation given is, in fact, a redundant statement. It B= 15
introduces x, y and z for unknowns, and at the same time retains R, B R = (4 x 9) = 36
and G. (In fact x = R, y = G and z =B.) G = (1 x 9) = 9
This is a basic error mathematicians often make - over -complicating
and so on, until we come to the last possible answer when there are
translation of facts into equations so that redundancies are introduced,
55 blue balls:
or the whole equation becomes too complex for easy solution. At the B =- 55
extreme, mathematical analysis can become so complex that only the R = (4 x 9) = 4
originator can understand the working - and as a result nobody can
G = (1 x 1) = 1

check the result!

Suppose, now, the information given in the problem is incomplete, There is no correct answer produced by this form of logic. There are
i.e. using the balls in a box example again, only two facts are given, not enough facts to establish one. In other words, on the information
the total of 60 balls and the number of red balls being four times the available, it is only possible to derive all possible answers.
number of green balls. There is no mathematical solution in this This method of deriving possible solutions by logic will also work
case. However, logic - and a little simple arithmetic - can give us when there are enough facts available for an exact solution, using only
the various numbers of different coloured balls possible within the some of the facts. The exact solution is then obtained by seeing which
facts. of the possible solutions fits the remaining fact(s). This is a more
lengthy process, of course, but in certain circumstances can be a
(i) The relationship between red balls and green balls is known simpler one when it seems difficult to assemble all the available facts
(R = 4G). Also there are three different colours of balls in the in equation form.
box, so there must be at least 1 blue ball. Real life problems, too, may not contain all the facts for exact
(ii) To be consistent with the fact that R = 4G means that the total solution. In that case, only possible solutions can be obtained. Numer-
of R and G together (R + G), must be a multiple of 5. (4 parts ical problems set as puzzles or 'mind -teasers' on the other hand invar-
for R plus 1 part for G.) iably contain all the facts for exact solution, although this does not
(iii) The largest available total is 60 -1= 59 (there must be at least
necessarily preclude the possibility of there being alternative possi-
1 blue ball). This is not divisible by 5. So the largest possible bilities.
total (R + G) is 55.
(iv) That means there must be at least 5 blue balls, leaving 55 to be
Illogical Logic
allocated between R and G on a 4 to basis, i.e. R = 4 x
1
Consider the following: Johnson said 'If my son's age was trebled
11 = 44 and G= 11. he would be as old as me. But in 15 years time he will be half my age'.
That is our first possible solution: B= 5 How old are we both now?'
R = 44
First let's solve this by logical arithmetic, using S for son's age and
G = II
J for Johnson's age. Assembling the equations from the facts:
(v) The next highest possible total is 50 (again divisible by 5). This
gives a second possible solution: Equation (i) 3 x S =J
Equation (ii) S+ 15 =-1 U+ 15)
B= 10
R. (4 x 10) =40 to make it simple, re -write this as
G=(1 x 10)=10
 60 Simple Arithmetic or Logic with Numbers 61
Logic Made Easy

Now since J = 3 x S from equation (i), substituting this in As a hypothetical example - and an exercise in logic - let's take the
equation (ii)
gives case of a printer who produces large catalogues for various clients.
2 x S+30 =3 x S+15 Trying to cut costs, individual clients often ask how much can they
Or 15 =S save by leaving off the page numbers? The printer's cost for setting
Thus if the son's age is 15, Johnson must be 45. these numbers is, say L5 per thousand digits. A particular client asks
what the savings would be on his 512 -page catalogue.
Suppose now the problem was a little more obscure, or we could The printer now sits down to think, and works things out like this:
not see how to derive the necessary two equations. In that case
try
solving the problem on an arithmetical trial -and -error basis, using a (i) Pages 1 to 9 require one digit for page numbering, making 9
little genuine logic and a lot of 'illogical logic'. digits.
If Johnson is talking about 15 years on, he is probably now under (ii) Pages from 10 to 99 require 2 digits for each number and thus
60. His son, too, will probably be 20-30 years 20 digits for each ten pages, requiring 9 x 20 =180 digits
younger than Johnson.
On this basis, let's assume that Johnson is 54, when his son's (iii) Pages from 100-999 require 3 digits for each number and thus
age must
be 18. 300 digits for each hundred pages. Four groups of hundred
What does this give in 15 years time? pages will take the book up to 499 pages, requiring 4 x
300 =1200 digits
Johnson will be 69 and his son 30. This does not agree with the second
fact - it shows a surplus of 9 years. So logically Johnson must be (iv) Pages 500-512 will then require 3 digits per page, on 13 pages
younger or 3 x 13 = 39 digits.
than 54 - and illogically
by that surplus. Thus Johnson is 54 -9 = 45,
and surprisingly that checks out as correct! Adding all this up he finds 1428 digits required, which at a cost of L5
Do not expect 'illogical logic' to provide correct answers every time. per thousand amounts to a sum of £7.14 if the page numbers are
It certainly won't. But it is the basis of another method of omitted. (His time in working it out will have been worth more than
handling
arithmetical logic. First an answer is estimated (or guesstimated, if that!)
you like) on a reasonably logical basis, i.e. what appears to be the Logically, when faced with a similar problem again he will simply
right order of answer. Then see how it fits the facts. From this, say to the client (from experience) that any saving is negligible. But
the original estimate up or down to get nearer to the facts,adjustuntil suppose the client persists in asking what the saving is for a particular
eventually your adjusted answer meets the facts. book. Does he go through the whole process again? Not if he has
What you are doing, indeed, is employing a sort of loose variation spotted the formula which can be derived from the original working.
of reductio ad absurdum
together with logical reasoning. The number of digits required fall into a distinct pattern:
Not to be recommended for general use - 'illogical logic' can become
pages 1-9 =1 x 9
tedious if dealing with two or more unknowns which have to be pages 10-99 = 2 x 90
estimated and continually readjusted. But when all else fails and
you pages 100-999 = 3 x 900
cannot see how to work out the arithmetical problem, you could try
it instead of giving up entirely! and so on (e.g. pages 1000-9999 for a really long book used = 4 x 9000
digits). So the basic formula is:
Devise Your Own Formulas
number of digits= (1 x 9) + (2 x90) + (3 x 900) + (4 x 9000)
Quite often at work or business similar tedious calculations pages covered 9 99 999 9999
crop up
at various times, which have to be worked out from first principles. A
desk calculator helps, but it still takes time. How very much simpler Any intermediate number of pages can then be inserted into this formula
if a working formula can be devised to cope with the problem. for a quick answer.
 62
Logic Made Easy Simple Arithmetic or Logic with Numbers
63
Lets use 512 pages again, which falls in the third
group. The fourth This could have been tackled in another
group does not apply and is ignored. Put the actual total of pages in way. The basic 'plan' for
their third group, and deduct the prior pages covered by the allocating notes into different pockets shows that a total of 28 notes
previous are required. So it was rather foolish of Bellamy to have spent all his
two groups:
money in buying $100 notes for with the amount he had available he
number of digits= (1 x 9) + (2 x 90) + 3 x (512 - 99) could only get 25 notes. He should have looked for the answer first,
=9+ 180+ 1239 and saved himself the trouble of going back and changing three
= 1428 notes.
He was also fortunate that there was a simple solution to his re-
Is it Possible? quirements.
Here is another example of a problem in arithmetical logic where it Jackson was much more methodical about planning his re-
is not known at the start whether there is a possible answer or not. quirements for his trip to America. He hit on the same basic plan, and
had the same amount of
money for buying dollar notes, but he went
Bellamy is travelling to America. He has bought twenty-five $100 about working it out this way.
notes from his bank. For safety he wants to distribute them in
different pockets, with each pocket holding a different number of (i) The least amount of notes I need are 28; and I can afford to
notes and different total value. He has seven pockets in all. Can he buy up to $2500 worth of notes.
do what he wants? (ii) I cannot afford to buy twenty-eight $100 notes
as that totals
$2800 - more than I have.
This problem is tackled from first principles. The least number of (iii) I will therefore buy twenty-eight $50 notes which will
notes he can have in any one pocket is 1; and the least difference in the cost me
$1400 and leave me $1100 still available
amounts in each pocket must also be 1 (in this case 1 note). So write (iv) I will then allocate these notes: 1 in pocket (i), 2 in
down numbers (i), (ii), (iii), etc, representing pockets and enter under etc. pocket (ii),
each the number of notes allocated on the above basis. (v) I can afford to
buy eleven more $100 notes. If I add one to
pockets (i) (ii) (iii) (iv) (v) (vi) (vii)
each pocket, that will still give me a different number of notes
notes 1 2 3 4 5 6 7 and different total value in each pocket. That will need seven
Add up the number of notes needed to do this. The total is 28 notes. more notes, costing $700 and leaving me with $400
unspent.
(vi) If I use this to buy four more $100 notes I can still conform
But Bellamy only has 25 notes, so he cannot fulfil his original plan. to
How, in fact, could he do it, for he cannot afford to buy more my original plan by putting all four of these additional notes
into either pocket (iv), (v), (vi) or (vii); or
dollars. Only changing one (or more) $100 notes into smaller denom- alternatively one
inations so that he can have a greater number of notes. extra $100 not into each of these four pockets.
The answer is quite simple in this case. Logical assessment of the Jackson has arrived at possible solutions by a mixture of
arithmetic and deductive logic. simple
position has established that he needs 28 notes, not 25. So Bellamy
changes three $100 notes into six $50 notes. That now gives him a Quite often problems of this type do not have an answer - see the
total of 28 notes, which he can now allocate like this: example under Reductio ad absurdum (Chapter 1).
pockets (i) (ii) (iii) (iv) (v) (vi) (vii) Looking for Short Cuts
number of notes 1 2 3 4 5 6 7
value of notes Here is what looks like. a tedious example of
$50 $50 $50 $100 $100 $100 $100 solving simultaneous
money in equations for unknowns A, B, C and D. A solution is possible
(assum-
ing all the equations are valid) because there are four unknown
pocket $50 $100 $150 $400 $500 $600 $700 four separate equations.
and
64 Logic Made Easy

(i) A+2B+3C+4D=2
(ii) 4A+4B+2C+2D= -2 CHAPTER 8
(iii) 5A + 2B + C - 3D = -6
(iv) 3A - 2B -C + 7D = 22
Let's look for a short cut. There is one obvious one we can try. Logic in Aptitude Tests
Adding equations (i) and (ii) together gives:
Aptitude tests are widely used in vocational guidance and industrial
5A + 6B + 5C + 6D = 2 -2 = 0
training. To be valid aptitude tests the questions must be solvable by
Now group as: a specific talent, not by knowledge or learning. Some, but only some,
can be solved by logic reasoning. Others have to be tackled by mental
5(A+C)+6(B+D)=0 reasoning, numerical reasoning or abstract reasoning. Still more are
Thus A + C =0, which means A = -C based on latent technological skills (fitting pegs into different hole
or C = -A shapes as an elementary example), but these are outside the scope of
and B + D = 0, which means B = -D a book on logic.
or D= -B An example of test questions which can be solved by logical reason-
ing is diagrams of a series of different shapes in different sizes and
Try substituting for C and D in equations (iii) and (iv) which then colours arranged in a pattern, but with some shapes missing. The
become:
problem is to find what the missing shapes should be. Since the
(iii) 5A + 2B- A + 3B= -6 which simplifies to: complete pattern is planned (by the designer of the test) on some
logical basis, there must be a logical answer.
4A + 5B = -6
The following is a very elementary test; the question to be solved
(iv) 3A - 2B + A - 7B = 22 which simplifies to: being what are the next three symbols in the sequence:
4A -9B = 22
We now have two simple simultaneous equations to deal with. Sub-
0- 0-0-0-0 -A- 0- 0-L-- ? -1 -? -
tracting (iv) from (iii) gives: Starting point is to identify the individual symbols by a simple code
which is easier to work with than shapes - say 1 for a square, 2 for a
14B= -28 circle and 3 for a triangle. The pattern is then rewritten in this
or B= -2, when D must equal 2
(number) code:
Substitute B= -2 in another equation, say (iii):
1 2 3 1 2 3 1 2 3 ? ? ?
4A- 10= -6
so A = 1, when C must equal - 1 It takes only a little study to see that the pattern is divided into groups
Smarten up your school algebra, if it has got a bit rusty. It can be of three which follow each other in the same manner, i.e. simply repeat
the quickest and simplest method of solving problems or puzzles in themselves:
logic involving numbers.
first group I
second group I third group I fourth group
1 -2 -3 I
1-2-3 I
1 -2 - 3 I
?-?-?
 66 Logic in Aptitude Tests
Logic Made Easy 67
Typically the fourth group must be the same as the others, variation that occurs is that the solid symbols appear consecutively,
i.e. fourth group moving one position to the left in each succeeding group. Hence we
would be (logically) justified in completing the third and fourth
1 -2-3 as:
groups
or re -write original symbol form
3rd group I (ii) 2 2 3
0- O -A 4th group I (ii) 2 2 3
These are, therefore, the missing symbols in the original Hence the missing symbols in the original are:
A little too simple and obvious? Then try to find the question.
in the following: missing symbols
last in third group
El A A 0 El A 0 III A A A - - - - - fourth group

First make a plan of the individual symbols involved.

These are
0 111 A 0
Square -- cell this 1.
This is a logical solution not necessarily completely correct, but the
Four versions of a triangle - piainL\ - call this 2 best we can do with the available information. Certainly the first
symbol in the third group is I or ; but the last symbol in the third
- solidA - call this 2 (it is still a triangle, group could be 3 or (there is no evidence in the original information
but different) to say whether it is likely to be or not). If it is 3, then by the 'shift to the
small triangle p - call this (ii) (' little 21) left' rule established, the fourth group could be either:
small solid - which is logically (T)
Circle - 0 -call this 3
1 (ii) 2 2 3 which is II AAA°
The original pattern is then re -written as follows: or 1 (ii) 2 23 which is 0L LAI
1 (ii) 2 2 3 1 (ii) 2 2 3 1(ii) 2 2
That, in fact, is the weakness of the technique of aptitude tests of this
This logically splits into groups of five symbols repeating themselves type. They are not necessarily capable of yielding only one answer by
with variations. To get a better picture, arrange these
groups in separate logical reasoning. There may be equally possible alternatives. Unfor-
lines:
tunately, when set as test questions there is only one correct answer -
1st group 2 2 3
the one the originator of the question has given.
1
(ii)
2nd group 1
(ii) 2 2 3
This ambiguity is even more possible in solutions to questions or
3rd group (ii) 2 2 problems set which have to be answered by abstract reasoning. Which,
4th group for example, is the odd one out in this series of diagrams?
We now have a picture of what is going on. All the
basic symbols
appear in the same order in each group, i.e. a square first, followed by 010 X °A 00
a small triangle, then (a)
two large triangles, and lastly a circle. The only (b) (c) (d)
68 Logic Made Easy Logic in Aptitude Tests 69

Reasoning would select (d) - the only one with the two circles on be anticipated as F = D + 50, G = E + 50, H = F + 50 and
one side of the line. But (a) is also an odd one out on the reasoning J=G+50).
that it is the only one with a vertical straight line. Also (c) is an odd Using the first relationship would give:
one out on the reasoning that it is the only one which is a mathematical
A B C D E F G H J
symbol (percentage). -
Surprisingly - or perhaps not so surprisingly - abstract reasoning (A- 10) (C- 10) (E - 10) (G - 10)
50 = 40 100 = 90 150 =140 ? ? ?
can be extremely puzzling, or even frustrating, to extremely intelligent
people who tend to think 'logically'. The number of extremely clever This gives the missing figures for F, but still leaving values for G, H
children who give anything but the right answer to the following and J unknown.
simple question is remarkable: Using the second relationship gives:
`If you were eating an apple and found half a maggot in it, what A B C D E F G H
would you think?' 2xA - 3xA 4xA - 5xA
Numerical problems are much simpler for logical minds. One just 50 40 100 90 140 -
150 200 250
looks for the pattern of 'formula' involved, in the same way as applying
logic to patterns of symbols. Again the classic problem is to find the This gives values for G and J. Now, knowing G, we can establish
missing or next number(s) in a series of given numbers which are from the first relationship that the missing value for H should be
known to have been generated in some logical way. What, for example G -10 =190. Thus the missing figures are:
are the next numbers which follow logically from: F G H J
90 150 ? 140 200 190 250
50 40 100 ? ? ?

To find the logical answer, write down the numbers as heads of Using the third relationship gives:
columns lettered A, B, C, etc and see what relationship there is A B C D E F G H J
between columns: A+50 B+50 C+50 D+50 E+50 F+50 G+50
50 40 100 90 150 140 200 190 250
A B C D E FGH
90 150 This confirms the solution derived from (i) and (ii). In fact, it
Number 50 40 100
- A-10 2xA A+B 3xA provides all the missing numbers, and so is a complete 'formula' for
Possible solution in itself. To check, see how the given values for B and A can
or or or or
relationships also be shown to be consistent with this formula.
D -A E/3 C-10 A+C
or B = number before A + 50 = 40
or or
A+50 B+50 C+50 Therefore number before A must be -10
A = number before the number before A + 50 = 50
Notice how three repeated relationships have appeared: Therefore the number before the number before A must be 0. In
other words, if the series were extended leftwards from A, it would
(i) a ' - 10' relationship in alternate columns B and D (which read:
could be anticipated as also following in columns F and H).
A
(ii) A 'multiplication' relationship in alternate columns C and E
(which could be anticipated as also following in columns G 0 - 10 50 40 100
check (0+50 (-10+50 (50+50
and J)
(iii) A `+50' relationship in columns C, D and E (which could also =50) =40) =100)
70
Logic Made Easy

Incidentally the preceding number value - 10 could also be de-

duced from relationship (i); but neither relationship (i) nor (ii) would CHAPTER 9
establish the number before that (0).
The validity of all the relationships has been proved by being re-
peated in the original information (the original series of numbers).
The relationship in the original column D (D = A + B) appears only
Introduction to Block Logic and
once, so there is no justification for it being applicable again in the Truth Tables
series. If it were, by the same logic:
since D=A +B solved by block logic using
Many problems in logic can be readily
then E should equal B + C or 40+ 100= 140, which it is not appropriate combinations of logic functions. The basic technique is very
(E= 150). simple for only three different functions are involved:
Thus this relationship is not repeatable, so is not valid. NO (normally called NOT), AND and OR
As a further example of numerical reasoning, find the missing The 'opposite working' or inverted forms of these are:
numbers in the following series: OR called
NOT NOT or YES; NOT AND called NAND; NOT
2 4 4 16 16 NOR
write down in column: The attraction of this method is that you can represent each logic
think of
function by a rectangle with its function written in, and then
A B C D E F G this block as a switch which accepts a premise or signal applied to one
side as an input and then either passes it through the block as an answer
2 4 4 16 16
A A2 A2 a truth
relationship or output signal, or stops it. At the same time we can construct
B2 B2 ? ?

The relationship pattern to emerge is that following pairs of numbers table which shows all possible combinations of signals in and out.
are equal (i.e. B = C and D = E, so F =G); and that such pairs are the A NOT logic block can accept only one input and has one output.
square of the preceding pair number. Hence F and G are 162 or 256. It works like a switch which is normally closed. Thus a signal applied
to the input opens the switch, sothere is no output - Fig. 9.1. Information
Signal on
Truth Table
Signal off
A S
A NOT 0 1

Switching equivalent 1 0

Fig. 9.1
shown by
fed to the block is rejected, i.e. NOT acceptable. This is also
the truth table as 'yes' or 'no' combinations. Conventionally 'yes' is
written as 1 and 'no' as 0 to avoid possible confusion with other logic
table.
blocks, e.g. a YES logic block has a different truth
The inverted form of NOT is YES, corresponding to a switch which
the
is normally open. On receipt of a signal the switch closes to pass
signal through as acceptable (YES) - Fig. 9.2.
 73
72 Introduction to Block Logic and Truth Tables
Logic Made Easy
Truth Table 51 Truth Table
Signal off A B S A S1 B S2 S
Signal on A A mi NOTH
Truth Table S 0 0 O 1 0 1 1
A YES NAND
A 0 0 0 0 1
II
S
e
1 B NOTI- 1

1 0 1
Switching equivalent 0 0 0 1 1
S2
O 1

1 1 1 1 0 1 0 1 0 0

Fig. 9.2 Fig. 9.5

Exactly the same result is obtained connecting two NOT devices in signals are present at both A and B is satisfied by there being no output.
series, when the output of the first becomes the input to the second - All the other combinations of A and B signals produce an output.
Fig. 9.3. As the truth table shows, the final output has the same Again note that the same (NAND) logic function is also performed by
relationship to the input A as YES (NOT NOT logically is the same an AND block with a NOT in each input line. (In this case calling for
as YES). However, this uses two logic blocks instead of one - or two three blocks instead of one.)
stages of analysis, if you like, instead of one. It is obviously simpler to Finally the OR block and its inverted form the NOR block are
of a
provide a YES function with a single block (single device). Neverthe- shown in Fig. 9.6. An OR block gives an output in the presence
less the inverted forms of the logic blocks are extremely useful. there is no
signal at A or B. A NOR block gives an output only when
Truth Table signal A or B (neither A nor B).
A
S1 A S1 S2 Signal off
S2 0 1 0 Truth Table
,Signal on ABS
1 0 1

O0 0
Fig. 9.3
OR H S 1 0 1

O 1 1

AND and OR blocks (and their inverted forms) again have a single 1 1 1

output but in this case can have two or more inputs - as many inputs Switching equivalent
as are necessary to accommodate the relative inputs in fact. For Fig. 9.6
simplicity, we will consider just two inputs being used.
Let's try a simple example of using logic blocks to solve a problem
Fig. 9.4 then shows an AND logic block with inputs A and B, and
in logic, taking one of the examples used in Chapter 2. (There is a
output at S, together with its truth table. There is an output (S= 1)
reason for this which will be explained later.) The example summary
only when inputs A and B both = 1 (we have used 'S' for output here,
is repeated here to save looking it up.
designating a signal). In this case there is a signal output when there
is an input at A and B. (i) Brown, Green and White are each wearing a coloured hat
(ii) The colours of the hats are brown, green and white
Truth Table
A 8 S (iii) Brown pointed out this fact
Signal off
on
Signal off
0 0 0 (iv) The one wearing the green hat then pointed out that none of
on
cotoolt_00002L__:%'. 1 0 0 the colours of the hats they were wearing was the same as their
B AND H Switching equivalent 0 1 0
names.
1 1 1

Fig. 9.4 Designate the hat colours B for brown, g for green and w for white.
Take Brown, as he spoke first and this must be of some significance in
The inverted form or NAND logic block is shown in Fig. 9.5. Here
the problem. Brown is wearing a brown hat or a green hat or a white
the truth table shows that when there is a signal input at A and at B
hat - so connect b, g and w as inputs to an OR logic block - Fig.
there is no output (S =0). Thus anything that is to be inhibited when
 V.

Introduction to Block Logic and Truth Tables 75

74 Logic Made Easy

b because they complement each other, when circuit designs for example
B
are worked out in mathematical logic, the final equation can then be
gh NOT
OR g OR B ---INOTH OR B
redrawn in block logic form to show the number and types of logic
devices required.
(a) Equally, if designed in block logic, the final diagram could then be
(b) (b) rendered in the equivalent mathematical logic equation which is then
Fag. 9.7
studied to see if it can be simplified.
The main use of both block logic and mathematical logic, in fact, is
9.7(a). Now from (iv), Brown cannot be wearing a green hat, so in functional circuit design - electrical circuits (ranging from quite
exclude this as a possible input by inserting a NOT in the g input line elementary switching circuits to those for microprocessors and com-
puters), and control circuits in hydraulics and pneumatics. Surpris-
- Fig. 9.7(b).
From (iv) again Brown cannot be wearing a brown hat, so exclude ingly, neither is used as much as it could be in solving `non-functional'
this as a possibility by inserting a NOT in the b input line - Fig. logic problems. Think about it before you automatically use deductive
9.7(c). logic for dealing in premises. If premises can be turned into 'signals'
The answer is now obvious from the final diagram alone - we don't operating on a yes -no basis (true or false), positive solutions can be
even have to bother with constructing a truth table. Brown can only obtained in terms of NOT, OR and AND logic.
be wearing a white hat. Equally, if you prefer to work with diagrams, then construct a truth
The reason for repeating this example is that block logic and mathe- table for the block logic involved. In many cases you can use the
matical logic (which was used for the simpler solution in the original simpler, more easily understood simple logic diagrams described in
example) are very closely related. Block logic, in fact, presents mathe- Chapter 5. The other alternative is Venn diagrams (Chapter 4),
matical logic in diagrammatic form since they cover the same logic although here there is a greater possibility of making mistakes until
functions and have the same truth table. One can be used to check the you become thoroughly familiar with this type of diagram.
other. If a question arises as to the validity of an equation in mathe-
matical logic (Boolean algebra), it can be 'spelt out' in the form of Block Logic and Circuit Design
block logic to see if it makes the proper sense. Block logic is a useful and readily mastered tool for circuit design
There is one important difference, though. The use of block logic, involving switching or control elements. For electrical circuits, each
especially for more complicated problems, can lead to redundant logic function AND, OR, NOT can be performed by a switch (or
blocks being introduced, i.e. more logic blocks than are strictly neces- combination of switches). In hydraulic or pneumatic control circuits,
sary. Also redundancies are not always easy to spot or eliminate. This similar functions can be performed by control valves. Circuits can thus
is not necessarily important using block logic for solutions in deductive be designed with logic blocks, and then finalized by substituting the
logic, but it is if block logic is being used to design control circuits. It appropriate switches or valves for each logic block.
means that the final circuit ends up by using more switching devices Let's start with a very simple example. Suppose it is required to
(logic blocks) than are strictly necessary. If the same problem is solved design a circuit to switch something on and off from two separate
in mathematical logic (Boolean algebra) it is readily possible to positions A and B. Drawn as a solution in block logic this simply
simplify equations and thus eliminate all redundancies. involves an OR function - Fig. 9.8. This is easily translated into two
Both block logic and Boolean algebra deal with logic in terms of separate switches, one at position A and one at position B, as in the
basic logic functions, and in an exact way (as opposed to drawing a second diagram.
conclusion from premises in deductive logic) - the one in a diagram- Suppose, now, it is necessary to be able to switch 'on' and 'off' from
matic way, and the other with mathematical equations. Equally, either station. The first circuit will not work. If switched on at A, for
 76 77
Logic Made Easy Introduction to Block Logic and Truth Tables

A Switching circuit expresses

the required logic in the form - either A OR B 'on' AND (at
A the same time) B OR A 'off'. In fact, we might have thought of
OR S planning the block logic diagram first in this way.
An interesting fact now emerges. The switching equivalent of this
block logic solution, shown in the right-hand diagram, is considerably
diagram (i)
more complicated than that for the alternative solution. In other
words the solution in OR logic is more complicated as regards switches
diagram (ii)
Fig. 9.8 than the solution in AND logic. The reason for this is that AND logic
operates in series fashion and OR logic works in parallel fashion. In
instance, and left on, it cannot be switched off at B, and vice versa. terms of mechanical switching, series working is less complicated than
Additional logic is required in this case, as shown in Fig. 9.9, with the parallel working. Circuit designs in AND logic, therefore, are nor-
corresponding switching circuit shown in the second diagram. Now if mally simpler than designs in OR logic when translated into mech-
switch A is left 'on', for example, switch B will be in its 'off' position anical switches. It does not follow, however, that this is also true when
and the circuit completed through its 'off' position (i.e. ri state). Moving the switching is done by electronic 'gates'.
As a further matter of interest, solutions to the above problem could
equally well have been derived from truth tables or Boolean algebra.
AND 1-L, The requirement, expressed in Boolean algebra is:
OR S
AR+AB=1
AND (i.e. A 'on' with B 'off' OR A 'off' with B 'on' gives an output). This
would translate in logic block form as in Fig. 9.9.
Fig. 9.9 Looking at the requirement the other way round, i.e. the combi-
nations that do not produce an output, corresponding to the complete
switch B to its 'on' position will then break the circuit (with switch A circuit being switched 'off'), the Boolean equation would be:
still 'on'), and thus switch off the circuit. The same occurs if switch B is
left 'on', with switch A 'off'. Switching A 'on' will switch the circuit AB+Aii=0
`off'. In fact the two switches do not have any specific 'on' or 'off' which for 'switch on' working has to be inverted and becomes:
positions. They are merely acting as 'gates' to perform a combination
of AND and OR functions. (AB+M3)=1,
It is also possible to produce the same working with a different which by applying de Morgan's theorem becomes:
arrangement of logic blocks, as shown in Fig. 9.10. This diagram
(A +B) . (A + r3) = 1
This would translate in block logic form as in Fig. 9.10.
OR
More Complex Circuits
AND I- S
The control problem solved in Fig. 9.11 by block logic is a little
irA
OR - more complicated. The device being controlled is a machine which
can be started from either of two operator positions A or B. Before the
Fig. 9.10 machine can be started, however, it is necessary that the component
 78 Introduction to Block Logic and Truth Tables 79
Logic Made Easy

Component
elements, alternative solutions are often worth looking at. Virtually
Component
A+B on machin sensor C any problem of this type can be solved either in AND logic or OR

--
Machine - Danger
logic; or the inverted forms NAND or NOR logic. In this case only
one type of logic element will be required, together with NOT elements
for essential inversion.
sensor D
Fig. 9.12 shows the same problem as Fig. 9.11 solved in NOR logic.
Danger area Specifically, this involves using two NOR devices and five NOT
devices - seven devices in all, but we have got everything in NOR

Fig. 9.11
it is handling is in position (as detected by the component sensor C),
and there is nobody standing in the danger area. Sensor D generates
a danger signal if anyone is in this area.
The logic required is A OR B AND C AND NOT -D. A and B are
C
connected to an OR block, and the output of this block provided one
input to the AND block. C is also connected to the AND block, giving D
a second input to this block. D is connected to the AND block via a
NOT block and is the third input to this block.
There will be an output from the AND block, switching the machine Fig. 9.12
on, only when there is an input signal at each of the three input lines to
the AND block. If either operator A or B switches the machine on, the
machine will only start if there is also an input signal on the C line logic. Also we can simplify this diagram. Two NOTs in the same line
(component in position) and no signal on the D line. The NOT in this simply change a signal back to what it was originally - they are
line will then invert this '0' signal into a 1', providing the third signal redundant elements which can be eliminated. On this basis we can
input to the AND block. If there is a signal on the D line (someone eliminate four devices from the original and end up with just three
standing in the danger area), the NOT in this line will invert that '1' devices performing the same function, still all in NOR logic - Fig.
into a '0' - i.e. there will be no input from this line into the AND 9.13.
block. Consequently there will be no output from the AND block and Try working out further alternative solutions in AND logic, OR
the machine will not start. In Boolean algebra this is presented quite logic and NAND logic only. (Remember you will have to use NOT
simply as: devices as well.)

(A +B) .C.b= 1
In practice - and with experience - circuit designs of this type are NOR
commonly developed in a specific type of logic. The solution in Fig.
9.11, for instance, uses a mixture of OR and AND logic elements -
NOT NOR
one of each in this case, but in a more complex control circuit there C
may be several of each required.
Depending on the availability of actual control elements, or possibly
looking for the simplest solution using the minimum number of logic Fig. 9.13
 V

80 Logic Made Easy

The main limitation with circuit design by block logic is that the
more complex the problem to be solved the greater the number of CHAPTER 10
logic blocks (devices) required and the more likely redundancies will
be introduced. Also it may not be easy to spot all the redundancies
and elimate them. Using Boolean algebra it is much easier to simplify
the equations. Even this has its limitations when it comes to the design Algorithms
of complex pneumatic and hydraulic circuits where Karnaugh maps An algorithm is a modern way of plotting logic problems (the term
can prove much more effective, but represent a new logic technology
to master. was not invented until the 1960s) which basically presents a chart
accommodating all the information relative to a problem and leads
The subject of circuit design by logic is further dealt with in Chap- result.
ters 11, 12 and 13. one exactly through the paths necessary to arrive at the logical
Basically, in fact, it is simply an extension of block logic, given a more
sophisticated name. It was originally devised as a means of plotting
the strategy of solving problems using computers. It is now widely
used for designing flow -charts, making it possible for people to arrive
with
at solutions on a yes/no basis (like the working of a computer),
the advantage that it can be completely non-mathematical. The scope
in this respect is considerable. Algorithms can be applied equally well
in general problem solving, commerce, industry, ergonomics, medi-
cine, finance, even political and military strategy. There are, equally,
mathematical algorithms. All have one thing in common. They are
directed to finding solutions - not teaching or learning. Algorithms do
not help understand a problem like a computer programme; therefore,
the solutions they give are only as good as the person designing the
complete algorithm.
Non -mathematical algorithms on individual facts or statements
relevant to the subject rendered as questions, which become in effect
`gates' in the flow path. Each 'gate' provides two exits - one for a 'yes'
answer and one for a 'no' answer, this procedure being followed
which is then the
through the algorithm until the final answer part (s),
logical solution to the answers made at the various question parts.
Rather than a single answer, too, an algorithm commonly contains a
number of different answers since it is designed to cover all contingen-
will make
cies, i.e. a 'yes' or a 'no' at all of the gates. A simple example
this clear:
A club has the following rules regarding subscriptions:
under
(i) Entrance fee is £50 for adult members; £20 for persons
20 years on January 1st.
 82 83
Logic Made Easy Algorithms

(ii) Annual subscription rates payable on September 1st each year case. This can save a lot of time and effort compared with working
are: out the answer each time for individual cases.
Full voting member The main problem to arise in designing algorithms - apart from
£50 ensuring that all the factors are included and presented on a yes/no
Senior (non voting) member £35 basis - is one of arrangement. Ideally - but not necessarily - all the
Junior (non voting) member £20 `yes' paths should come out the same way, and all the `no' paths in the
An algorithm drawn up on this basis would look like this: same (but different) way. Thus in the example drawn, a 'yes' auto-
Payable
matically reads straight across to the right. A `no' reads downwards
START New YES Over YES Voting YES and then across on a different path.
HERE £100
member 20? member The other difficulty, particularly when a large number of different
NO NO NO `boxes' have to be used is to avoid crossing paths, as this can be
£85 misleading. Also the final 'answering' boxes should be made distinct
from the others (e.g. with a bolder outline).
£ Drawing up the basis of an algorithm is quite a simple exercise in
Over YES
block logic once you have got the facts broken down into simple
Voting YES
20? member £ questions. Planning and drawing up the final algorithm, however, can
be something of an art to get it into its best presentable form. It may
NO NO
£35 even be necessary if the algorithm tends to become excessively com-
plicated to stop at certain answer part(s) which, instead Of being
Ic2oI answers, are new starting points for a separate algorithm.
This chart plots all the possibilities with logical paths through to
Algorithms and Truth Tables
the amount payable. It can provide immediate solutions relative to any
An algorithm can also be presented in the form of a truth table, from
applicant, new or existing member, whether over or under 20, and which relevant solutions can be drawn. To demonstrate this we will
whether voting or not. In the same way it could be extended to cover
other club subscription rules. Try re -drawing it to take into account
use the same example as before, dealing with club subscriptions,
as well: writing out each question as a line and covering all possible combi-
nations of answers in the following columns of the table:
(iii) Members joining after March 31st in any year pay only half
2 3 4 5 6 7 8
that year's subscription. 1

New member? YES YES YES NO NO NO YES NO

(iv) Family membership (two persons from the same family) can
have 25 per cent reduction on subscription rate. Over 20? YES YES NO NO NO YES NO YES
(v) Existing members over 65 years old, with at least five years Voting member YES NO NO NO YES YES YES NO
paid -up membership, can have a reduced membership of £40 Voting £100
for full membership, or £20 for individual (non -voting) mem- £85
bership.
£50 \/
Algorithms of this type can readily be prepared to cover almost all £40 N./
types of subjects involving variable factors which can be dealt with on N./
a yes/no basis. Once prepared, they then cover all possibilities involv-
£35
ing these factors, with a logical route to the correct answer in each £20 1 N//
84 Logic Made Easy

The truth table is more explicit in that it analyses all the possible
combinations of the variable factors (answers to the questions). There CHAPTER 11
are three questions, each with two possible answers (`yes' or `no'),
which in fact gives us eight possible combinations, as given by the eight
columns in the truth table. Introduction to Boolean Algebra
In this particular case the combination in columns 5 and 7 cannot in a mathematical
apply by the club rules since the person is under 20 and thus a junior The simplest approach to solving logic problems
which is certainly not as frightening
member, and a junior member is a non -voting member. With a way is the use of Boolean algebra, in many ways, but certainly
different set of questions all possible combinations may be relevant, in as it sounds. It is like ordinary algebra
much simpler because it is concerned with only two possible states of
which case there would be eight separate solutions, not six as in this false, yes or
example. Plotting a truth table, in fact, is a good cross-check on the each individual subject. These can be evaluated as true or
premises); or on or off, go or
completeness of the algorithm. no, in the case of general subjects (e.g. is 'I'
stop, in the case of signals. (The usual interpretation with signals
for 'on' and '0' for off'). because
At first it may look more complicated than ordinary algebra
for AND; + for
the mathematical signs are used in a different sense,
and A + B means A or
OR. Thus A B means A and B (not A times B);
B (not A plus B). inversion
The only other symbol used is a -over a letter, meaning an
means not B, and so on.
or logic NOT. Thus A means not A, in terms of
It then becomes obvious that Boolean algebra works NAND and
NOT, AND and OR logic, or the inverted forms YES,
NOR. is to think
By far the simplest way of understanding Boolean algebra
for logic
in terms of switching circuits applied to the basic equations
functions, as in Chapter 9.
the possibility
NOT then becomes a normally closed on -off switch within
of either a negated or positive output. The two Boolean algebraic
forms of NOT are thus:
A S
1 0
A=S
consistent with the truth table
0 1
A=S
The two Boolean
YES then becomes a normally open on -off switch.
algebraic forms of YES are thus: S A
1 1
A=S
consistent with the truth table
0 0
 87
86 Logic Made Easy Iritroduction to Boolean Algebra

AND is the equivalent of two normally open switches in series. Here the another form of OR known as an EXCLUSIVE
There is, however, when
presence of an input signal at either A or B closes its respective switch OR. This excludes the possibility of an output being generated
to pass on the signal to the output, but there is no output until both at both inputs. That is, the production of an output
signals are present
switches are closed (signals A and B are present): is exclusive to either A or B input signals being present.
NOR again is consistent with parallel switches giving an output
A B=S consistent with the truth table A B S when there is not a signal input at A or B. This
(both switches closed) will both
0 0 0 time either signal A or signal B (or both together) open
1 0 0 switches, giving no output.
0 1 0 A+B=S consistent with the truth table A B S
0 0 1

1 1 1
1 0 0
NAND is not quite so easy to describe in simple switching terms. 0 1 0
Logically it should be two switches in series operating in the inverted
1 1 0
mode to AND. In fact it works in the manner of one normally closed
switch which needs signal inputs at both A and B to open it. With no with the exception of NOT and YES which can only
Note here that,
input at either A or B the switch remains closed, providing a through all of the logic functions can accom-
have one (input) designation, which
path to the output and an output signal (consider this signal as coming A,B,C,D, etc, as necessary
modate any number of input signals
from a separate source). With an input at A the switch still remains of Boolean algebra. These merely extend
can be expressed in terms here is the
closed, so there is still an output. Similarly with an input at B only. the size of the truth table in the same truth sense. For example,
With inputs at both A and B the switch is opened, breaking the circuit accommodate four inputs A, B, C, D.
AND truth table extended to
and changing the output to 0.

A B= S
A B
0
CDS
0 0 0
consistent with the truth table A B S 0
0 0 1
0 0 0 1 0
1 0 1
0 0 1 0 0
0 1 1
0 0 1 1 0
1 1 0 0 1 0 0 0
OR is straightforward again - two normally open switches in parallel. 0 1 0 1 0
There is an output when either switch A or switch B is closed (by signal 0 0
0 1 1

input at A or B, respectively. Also there is obviously an output when

0 1 0
both switches are closed, as the truth tables shows: 1 1

1 0 0 0 0
A+ B= S consistent with the truth table A B S
1 0 0 1 0
0 0 0
1 0 1 0 0
1 0 1

1 0 1 1 0
0 1 1

1 1 0 0 0
1 1 1
 89
Introduction to Boolean Algebra
88 Logic Made Easy
be
A B CDS We still have the output inverted (5), which may or may not
convenient. If not, we can invert the whole expression again:
1 1 0 1 0
1 1 1 0 0 when A + B= S becomes A + B = S
1 I 1 1 1
Thus we started with not -A and not -B = S (or NOR in AND logic)
and
end up with not -A or B = S (or NOR in OR logic, which was what we
This illustrates the basic rule that for N inputs there are 2" possible
states or combinations, represented by N lines in a truth table. Thus were after). When we
in this case there are sixteen lines (different states) possible with the At this point there is something very important to note.
switching equivalent of sixteen on - off switches connected in series. first inverted the equation we changed the sign from (AND) to +
which demands this,
There is an output (S = 1) only when all sixteen switches are closed (OR), according to the rule of Boolean algebra
(i.e. A= 1, B= 1, C= 1 and D= 1). The truth table represented this i.e. theorem (xii). When we did the last inversion, we did not change
be
more compactly than drawing out all the individual switches con- the sign. This is because we inverted A + B as a whole, which can
considered as putting it in brackets (A +B) and treating it as a single
nected together. theorems
term. Again this is consistent with theorem (xii). (The
Inverting the Function governing Boolean algebra are summarized together on pages 94-5
One thing may - or may not - have become apparent from the logic for convenience of reference.)
equations explained above: inverting a function changes the nature of
the function. Thus AND which is obviously an AND function (A B or Ultra -Simplified Symbolic Logic
A and B), becomes an OR function when inverted to NAND (not A or Now let us look at an ultra -simple symbolic method
-
of handling
borrowed from
not B). This is a very important rule in manipulating Boolean equa- deductive logic, in which we use just the inversion sign
tions, where it is often convenient, or necessary, to invert functions to letter symbols. It is not Boolean alge-
Boolean algebra, together with but it
obtain an optimum solution (eliminate redundancies, for example). bra, and indeed the validity of the method is questionable -
Specifically, too, it enables all expressions in an equation to be in the works. Take the following as an example:
same type of logic.
Suppose we want to work entirely in OR logic. This means that all Premises: all dogs are animals
the terms used in an equation need to be expressed in the form A + B, no animals can fly
i.e. the + (OR) sign is used throughout the equation. Conclusion, by deductive logic: no dogs can fly.
Starting with the NOR function, A f3 = S , this is in AND logic (as F
shown by the sign). By the rule mentioned above, inverting it will Symbolic technique: designate all dogs D, animals A, and fly
change it into OR logic: first premise becomes D=A

thus A ft = S becomes A + B = S by inverting each term and second premise becomes A = F

Therefore if A=F
changing the sign A=
This now contains double inversions on A and B, which is bad logic. thus D=A=P
If inverted once (in the original form) and then inverted again (in the
conversion to OR logic), any subject ends up the right way up - or in (All dogs cannot fly, which is the same as no dogs can fly.) Try this
technique out for yourself using the examples in the chapter on De-
this case A = A and B = B, so NOR converted to OR logic now becomes:
used this quick
ductive Logic (Chapter 2). In fact, we have already
A+A=A+B-=S
90 Logic Made Easy Introduction to Boolean Algebra 91

method to provide a second solution to the 'coloured hats' problem in (iv) Since I has no other value than 1 (positive)
Chapter 2. Let's see how this same solution can be worked out cor- A+ 1 = 1 (OR function)
rectly in Boolean algebra, using the same notation as before - B for (v) A A = A (AND function)
Brown and b, g and w for the brown, green and white hats respectively. (vi) A + A = A (OR function)
The start is exactly the same as before: (vii) A A=o
(viii) A+A=1
B=b+g+w (ix) A=A (double inversion returns the function to its ori-
also because the man who spoke after Brown was wearing a green hat, ginal state)
Brown cannot be wearing a green hat. Nor can Brown be wearing a (x) The order of individual functions is immaterial
hat the same colour as his name, so AB=BA
A+B=B+A
B= g 6 (xi) The order of grouping of functions is immaterial
A. (BC) = (AB)C
Last time we jumped straight to the answer by assuming that g and 6
cancelled out g and b in the first equation. This is not justifiable in
A+(B+C)=(A+B)+C
(xii) Inversion of all signals changes the sign (de Morgan's theo-
Boolean algebra since there are not two separate equations involved,
rem)
only one combining the two propositions, viz: ABC=A+$+C
B=b+g+wandkS or A+B+C=A13C
(xiii) For simplification of equations . (AND) is treated as a multi-
B= (b +g+ w) gb which expands to
B = b 6.g + ggb +wg6 and then groups in OR logic as plication sign and + (OR) as an addition sign, as in ordinary
B=(bSg)+(gg6)+ (wgG) algebra
From theorem (i) b 6=o A(B+C)=AB+AC
gg =0 (xiv) Expansion of equations treated as in ordinary algebra
thus B= (0 g) + (0 b) + w g b (A+B)(C+D)=AC+AD+BC+BD
From theorem (i) 0 g=o (xv) Redundancies can be eliminated, e.g.
0 6=o A+ (AB)=A
Hence B=wgb A+ (AB)=A+B
=white, NOT green, NOT brown. This last theorem may not be obvious, but can be proved by its truth
table:
Exactly the same answer as the short cut solution gave!
A B S
0 0 0
Boolean Algebra Theorems i 0 1

(i) Since 0 has no value other than zero

1 1 1
A 0 = 0 (AND function)
(ii) Since 1 has no other value than 1 (positive) 0 1 1

A 1=A (AND function)

(iii) Since 0 has no other value than zero This shows that A =1 makes S= 1 OR B= 1 makes S= 1 whether
A+ 0=A (OR function) A= 1 or A=0. Thus A is redundant in the equation.

A.
 Introduction to Boolean Algebra 93
92 Logic Made Easy
Repeating this table but this time using only minterms and adding
Minterms and Maxterms
Boolean algebra and truth tables are closely related, as we have the output:
seen. One can always be expressed in terms of the other. However the A B C minterm output (S)
logic involved may be a mixture of AND, OR and inversions of AND 0 0 o A$e
and OR. Assuming that we start with a truth table, for a systematic 1 0 0 ABC 1

approach to combinational logic the truth table needs to be expressed

in a Boolean expression with a particular form, i.e. AND logic or OR
o 1 o ABC
logic. In AND logic the Boolean expressions are called minterms; and 1 1 0 ABC
in OR logic the Boolean expressions are called maxterms. The two are o 0 1 A$C
complementary, as solutions can be worked in either. Here, for ex-
0 A 11c
ample, is the truth table for covering three binary variables A, B and
1 1

C and the corresponding Boolean minterms and maxterms: o 1 1 ABC 0

A B C minterm maxterm 1 1 1 ABC 1

o 0 o ABe A -143-1-e C, A B C, A
Here it is seen that there is an output ifany ofA C
1 0 0 A.B.e A+11+e and A .B.0 is 1. The corresponding Boolean equation is then:
o 1 0 ABL A+B+C S=A11C+ABe+AclC+ABC
1 1 0 A BC A+B+O
A
0 0 1 A$C A+ii+c NOT AND
B
1 0 1 A tC A +11+c
NOT
o 1 1 ABC A+ B + C
1 1 1 ABC A+B+C
This is a particularly useful technique for the design of combina- B AND
tional circuits (e.g. electrical circuits or hydraulic or pneumatic con- NOT
trol circuits). If the relationship (i.e. significance of the 'switching' OR
points A, B and C) are stated in minterms, the circuit can be designed
using digital devices having an AND function, plus NOT for inver- A NOT
sions. If the relationship is in maxterms, then the circuit can be B AND
-i NOT
designed using OR devices, plus NOT for inversions.

B
 N

94 Logic Made Easy

Fig. 11.1 shows this as a complete circuit, using logic elements. It is an

unnecessarily complicated circuit for it contains a number of redun-
dancies. The original equation is capable of simplification (using the CHAPTER 12
relevant Boolean algebra theorems) to
S=AB+AC+ABC More Logic Equations
This gives the simple circuit of Fig. 11.2.
One particularly useful feature of Boolean algebra. is that it is possible
to express all the logic functions involved in a single logic. Problems
AND involving OR and AND, for instance, can be expressed entirely in
either OR logic or AND logic - or NOR or NAND logic. Working
out how this can be done is also useful practise in manipulating
Boolean equations and applying the rules laid down in the theorems.
AND OR
C NOT Everything in AND Logic
In AND logic the basic expression is A B = S. Thus to be in AND
logic every piece of logic must be expressed in the form.
An OR function is changed into AND logic as follows. Start by
writing down the basic OR function:
B AND
A+B=S
invert AI1=S
Fig. 11.2 Note this changes the sign from OR to AND, but the output is
inverted. Invert again to obtain a positive output:
Or as a 'wiring' diagram, as in Fig. 11.3
AB=S=S
Thus Af3=s is the OR function in AND logic.
AND
The NOR function is an inverted OR, and so is already in AND
form with a positive output. No conversion is necessary:
As=S
AND OR
The basic NAND function is in OR logic, viz:
C 4-1 NOT A -f -B=s
invert to give A B =
invert again for a positive output AB = S = S
AND Thus AB= S is the NAND functions in AND logic.

Everything in OR Logic
Fig. 11.3 In this case every piece of logic needs to be expressed in + form.
 w
96 Logic Made Easy More Logic Equations 97

Starting with the AND function: invert to give

A B=S A11=S
invert to put into OR logic: invert again to get the form
A+D=S All= S=S
invert again for a positive output: Thus AA=S is the OR function in NAND logic.
A+11=S=s The NOR function is A11=S
invert as a whole into the form
Thus A+11=S is the AND function in OR logic.
The NAND function is already in OR logic, i.e. At=S
invert again to get a positive output
A+11=S
The NOR function is in AND logic, i.e.
At=S=s
Thus At=S is the NOR function in NAND logic.
At=s
invert to put into OR logic Evegthing in .NOR Logic _
This time everything needs to be expressed in + form.
Ai -B=S The AND function is A B = S
which is the same as A +B= S invert to give
invert again to get a positive output A+t=S
A -I-B = S invert again to get into + form
Thus is the NOR function in OR logic. A+f3=g=S

Everything in NAND Logic

Thus A+S=S is the OR function in NOR logic
In this case since NAND is inverted AND, every piece of logic needs The NAND function is A+t=S
to be expressed in-form. invert as a whole to get into + form
Starting with the AND function A+B=S
AB=S invert again to get a positive output
invert to get the NAND form A-43=g=S
AB=S Thus A+B=s is the NAND function in OR logic.
invert again to get a positive output Circuit Design by Equations
AB=S=S Examples of circuit design by block logic have already been given
in Chapter 9 where, too, Boolean algebraic equations were used to
This is the AND function in NAND logic. check the results. (You probably skipped that part if you had not
The OR function is A +B=S already read Chapter 11.) All combinational control circuits can be

A
 98 99
Logic Made Easy More Logic Equations

expressed in terms of block logic or pure equations - or the re- In terms of logic we have the following requirement (or logic equa-
quirements equally well written out in the form of a truth table (which, tion):
of course, will be the same for the block logic and algebraic solutions). A OR B AND C NAND D AND NOT E = S (ignition signalled)
Provided you are happy working with Boolean algebra, solution by
equation is usually simpler and quicker, and redundancies can be which in symbols is:
eliminated in the process. A+BCDE=S
All combinational control circuits are similar in that they provide
This is in mixed logic (OR and AND logic, together with NOT). A
an 'on' signal when, and only when, a particular combination of other
control circuit, based on logic elements, would be as in Fig. 12.2.
signals are present. Thus the example given here is basically the same
sort of problem as solved by block logic in Chapter 9, but a little more A A+13
complicated in involving a combination of five different signals. Cir- OR
B
cuit design will be worked out in Boolean algebra on a logic basis, but A+BCDE = S
C AND
also illustrated by block logic diagrams as 'pictorial proof' of what the
equations mean.
The control problem, shown in Fig. 12.1, is this. A space rocket site
incorporates the following features. At the control centre, either tech-
Ej
D

NOT

Fig. 12.2
nician A or B can give the final signal to start ignition. The rocket is
fuelled up, ready to go, but it must be impossible for A or B to launch It could equally well be rendered in terms of AND devices only
the rocket until all the monitoring signals are in a 'go' state. These (AND logic), together with NOT.
monitoring signals are derived as follows. Starting with the original equation:
A+BCDE=S
invert once
Command centre Emergency sensor E
A.A+C+15+i.5
Display
E
signal D invert again to get into AND form and also give a positive output:
IA
2 :

---1 A 11CD E=S=S

This could be rendered in logic element form as shown in Fig. 12.3,
- - --------------
Ignition signal Control sensor C
Fig. 12.1
=-= = =_
A-E.C.D-E = S
A control sensor (call it C) gives a continuous signal indicating that
the rocket is sitting properly, ready to go. A display screen (call it D)
monitors the internal systems, generating a ready -to -go signal when
all functions are operating correctly. A further emergency sensor (call
it E) on the tower only gives a signal if something else is wrong and
the launch must not take place. Fig. 12 3
 100 101
Logic Made Easy More Logic Equations

but with obvious redundancies (NOT devices following each other for
double inversions and effectively cancelling each other out). NOR H
This is also obvious from the equation. Double inversions cancel A+B+6+5+E = S
NOT NOR
each other out, so the simplest form of the equation in AND logic
becomes (by remaining double inversions) D NOTI
A11CDE=S E

The requirements for this circuit are shown in Fig. 12.4. It has saved Fig. 12.5
six (NOT) devices.
Incidentally a truth table written out for this particular requirement
would run to 32 lines, only three of which would give a 'go' output
NOT H. (S=1)
A

B NOT I...,
AND H NOT AB CDES
line 1 ... 0 0 0 0 0 0
C AND 7aEcoE = S
1 0 0 0 0 0

NOT

Fig. 12.4

Perhaps it could be made even simpler using another form of logic? 1 0 1 1 0 1

Certainly it can. Suppose we try NOR logic. This means reading the
terms of the equation in the form +
Starting with the original equation:
A +BCDE=S
invert, treating A + B as a whole to retain the + form: 0 1 1 1 0 1

A+B+C+1)+E=S
invert again as a whole to retain the + form and get a positive output:
A+B+e+C)+E=S=S
Now remove all the redundancies (double inversions) - in this case
1 1 1 1 0 1

there is only one (E):

line 32 ... 1 1 1 1 1 0
A+B+C+17)+E=S
Even Boolean algebra, however, has distinct limitations when it
This can now be rendered in circuit form with just four logic elements,
as in Fig. 12.5. comes to more complete circuit designs. It can produce cumbersome
 'If

102 Logic Made Easy

equations which can prove tedious to work with, and need careful
checking for possible errors, especially when simplifying or trying to
convert into a different form of logic. Other types of logic are generally CHAPTER 13
to be preferred in such cases.

Karnaugh Maps and Logic Circuits

A Karnaugh map is a most useful device for combinational circuit
design (and can also be used for working other solutions in Boolean
algebra). It consists of a square or rectangular area divided into
The number of
squares, each square representing one minterm.
squares required will be 2", where N is the number of variables. the
Taking just two variables A and B as the simplified example,
corresponding Karnaugh map will have 22=4 squares, covering the
four possible combinations of A and B (i.e. A il, A B, A 11, A B).
These squares can be labelled in various ways, two basic methods
being shown in Fig. 13.1. That on the left labels the possible states of

,4 A B\A
0 1

If3 AB AB 0 AB AB
B 413 AB 1 AB AB

Fig. 13.1

A and B in columns and squares, respectively. That on the left desig-

nates A and B allocation - to the right and down from the separating
line at the top left corner, with possible signal values. The resulting
combinations marked in the squares are shown in minterms, but
alternatively could have signal values (i.e. binary 0 or 1). Of the two
the left-hand form is probably easiest to use for logic expression reduc-
tion (i.e. elimination of redundancies or minimization); and the right-
hand form probably easiest for translating a truth table into a Kar-
naugh map.
The sequence in which the variables are presented is not significant,
but the squares cannot be allocated at random. They must be
arranged so that movement of one square upwards or downwards, or
one square left or right horizontally, results in the minterms associated

Ail
 105
104 Logic Made Easy Karnaugh Maps and Logic Circuits

with the two adjacent squares differing only in a single variable.

1 reaching the edge of the map, the next move of one square
returns to
Kar-
Specifically this implies that the minterms in adjacent squares are the opposite edge of that square or column. In other words, a
in two
identical except for one variable which is marked in one square but naugh map is basically a three-dimensional device, drawn
not in the other. dimensions. How it continues in the third dimension can be under-
vertical
Karnaugh maps can accommodate more than two variables. Figs. stood by thinking of the flat map being rolled up into a
13.2 and 13.3 show alternative versions of maps for three and four horizontal movement), or a horizontal cylin-
cylinder (for continued
variables, respectively. This is about the practical limit. Maps with der (for continued column movement) - Fig. 13.4.
more than five variables are too cumbersome to be of real value.

A
00 01 I 11 10

o ABC ABC ABC ABC

C 1 ABC ABC ABC ABC

Read around map as If continuous
5 B Fig. 13.4
Fig. 13.2
Minimization
Movement in a Karnaugh map must always be horizontally (along A particular application of a Karnaugh map is for producing a
squares) or vertically (up and down columns), one square at a time. On minimal form of Boolean equation. We will use the original equation
used in describing minterms as an example:
A S=A$C+ABO+ABC+ABC
13.6
A8 A e AB A E
The Karnaugh map drawn for this is shown in Fig. 13.5. Fig.
C8 58 55 o5 then shows this map redrawn and re -labelled using the circuit outputs
5 A A
A6 AB AB AE
SD co -6 D -6 D
f 11
1

AE AB AB Al -3- ABC ABC ABC ABC

CD CD CD CD
C ABC Ci
AE AB AB AE ABC ABC ABC
C 5 c5 c5 c5 D
B B B B B B
Fig. 13.3 Fig. 13.5
 107
106 Karnaugh Maps and Logic Circuits
Logic Made Easy

A A
r if
- 1
0 1

110 1
i

0 0 r1, 1 NI
C
r 1 11 0 0
;
r 1

:1 1i 0 0
,

0 ) 0 C
0 (1 11 0 1 1

B B B
Fig 13.6 Fig. 13.7

of
relative to the various combinations. (With practice this could be done A basic problem in conforming to these rules is that the effect
13.7. Here
without going through the intermediate stage of Fig. 13.5.) `running off the edge' of the map is overlooked, e.g. see Fig.
smaller
Simplification or minimizing then consists of grouping together ad- there is already one large group of four squares, and two
bottom. In fact these two
jacent squares which both contain 1. Grouped squares can then be groups each of two squares at the top and
represented by a single AND term instead of two. There are two such smaller groups form one larger group (think of the map being rolled
groups in Fig. 13.6, shown enclosed by dashed lines. In the horizontal up into a horizontal cylinder). There are, in fact just two groups not
group, B is the variable that changes and so instead ofA BC+ASC three, needing just two AND terms, not three.
which this group covers, this group can be represented simply by A C Failure to spot how small groups may form a larger group does not
(i.e. A B C + A ft C in the original equation becomes A C (B+11) matter greatly. It does not upset the validity of the treatment. It
as
which = A C). simply means that the final answer derived is not as fully simplified
Similarly in the second (vertical) group, A B C + AB C, C is the possible. Translated into hardware in a switching circuit, for example,
variable that changes, so these two squares can be represented by a the circuit would work just as well but include more switching or logic
single end term A.B. There are no other groups and so the original components than are strictly necessary.
equation:
Sequential Circuits
S=A13C+ABC+ASC+ABC So far we have only dealt with combinational circuits. Other control
reduces to circuits may require sequential logic, i.e. operations or movements
the
S= AC +AB+ ABC following in a particular sequence. Even more complicated is
compound circuit which incorporates both combinational and sequen-
Basic rules for minimizing are: tial requirements.
All the squares containing a 1 must be included in at least one Karnaugh maps can again be used for the design of sequential or
(i) where sequence is involved a further
group (if not possible, there is no possibility of minimization). compound logic circuits, only
the right timing
(ii) Form the largest possible groups. type of logic is required - MEMORY for providing
of the sequence. One memory will be required at each sequence point,
(iii) Produce the smallest possible number of groups (consistent with
when each memory will require its own
Karnaugh map. In a com-
(i) and (ii) above.
 109
Karnaugh Maps and Logic Circuits
108 Logic Made Easy
B and not A = S
pound circuit, a further Karnaugh map (or maps) would be necessary or 13- A=S
for the combinational requirements. BA =0
Dealing with a typical sequential logic circuit, the procedure would the signal path
be as follows: In other words, when a signal is present at A it inhibits
of B. two
MEMORY is readily performed by a flip-flop device accepting
(i) Write down a word statement of the sequence. A will give an output
(ii) Draw a time diagram designating the complete sequence. signal inputs A and B. A signal applied when
at input
signal A is removed. A
state S1 and maintain this state even
(iii) Prepare a separate signal flow diagram and on this plot the of the memory to S2, which
signal input at B will then change the state
signal flow path. until tripped back
it will continue to hold when signal B is removed,
(iv) Prepare a Karnaugh map for every memory, and for every A. In other words,
to state S1 when a further signal is applied to which
auxiliary memory appearing in the signal flow diagram. MEMORY is a logic device which 'remembers' the last signal,
(v) Extend the presence of memory designations by infilling and
is an essential feature in sequential logic. the available
forming loops on the signal flow diagram.
As far as logic circuit components are concerned (i.e.
(vi) Finally, minimize the memory set and reset equations. circuit), choice
`hardware' for turning a circuit design into a working
on the suitability of such 'working'
of type of logic used depends This is no real
No wonder this form of logic is largely incomprehensible to all but elements to perform the various functions necessary.
specialists! solid state devices, but in the case
problem in electronic circuits using elements involved
Even Karnaugh maps have their distinct limitations when it comes of pneumatic and hydraulic circuits, the working
to the design of the more complex logic control circuit where, in fact, of individual
are normally miniaturized valves. If a multiplicityfavours using NOR or
only design by logic can be relied upon to provide optimal solutions. elements (valves) is to be avoided, the choice
For that reason many major producers of logic control devices them- simpler valves.
NAND logic, as these functions can be performed by
selves (hardware) have developed their own system methods to make -functional devices based
Other systems may, however, employ multi
logic designs easier to understand and minimize circuit design time. on OR, AND, NOT, YES and MEMORY.
This will then relate to a specific system which, in turn, is based on a
particular choice of logic elements providing all the logic functions
necessary.
For a full coverage of all the types of logic required in complex
circuits design the functions required are:

NOT AND OR YES NAND NOR INHIBITION

and MEMORY.

That does not mean that all these separate functional elements will be
required. Solutions can be worked all in AND logic, or all in OR logic
(or the inverted forms NAND and NOR). The only functions which
remain necessarily common with any type of logic chosen are NOT,
INHIBIT and MEMORY. NOT we already know about. INHIBI-
TION is a sort of variation on NOT and AND, expressed as:
 Help Us Help You
Other Bestsellers From TAB So that we can better provide you with the practical infor-
mation you need, please take a moment to complete and
0 SCIENCE FAIR: Developing 0 PARTICLES IN NATURE: return this card.
a Successful and Fun Project- THE CHRONOLOGICAL DIS-
Maxine Haren Iritz, Photographs COVERY OF THE NEW 1. I am interested in books on the following subjects:
by A. Frank Iritz PHYSICS-John H. Mauldin CI architecture & design electronics
Here's all the step-by-step guid- If you're interested in physics, O automotive D engineering
ance parents and teachers need to science, astronomy, or natural his- aviation 0 hobbies & crafts
help students complete prize -quality tory, you will find this presentation El business & finance 0 how-to, do-it-yourself
science fair projects! This book pro- of the particle view of nature fas- El computer, mini & mainframe military history
vides easy -to -follow advice on every cinating, informative, and entertain- O computer, micros 0 nautical
step of science fair project prepara- ing. John Mauldin has done what o other
tion from choosing a topic and defin- few other science writers have been
ing the problem to settling up and able to accomplish ... he's reduced 2. I own/use a computer:
conducting the experiment, drawing the complex concepts of particle Apple/Macintosh
conclusions, and setting up the fair physics to understandable terms fl Commodore
display. 96 pp., 83 illus., 8 1/2" x and ideas. This enlightening guide CI IBM
11". makes particle physics seem less El Other
Paper $9.95 Hard $16.95 abstract-it shows significant
Book No. 2936 spinoffs that have results from re- 3. This card came from TAB book (no. or title):
search done, and gives a glimpse of
future research that promises to be
of practical value to everyone. 304
pp., 169 illus. 16 Full -Color Pages,
4. I purchase books from/by:
0 general bookstores 0 telephone
14 Pages of Black & White Photos.
technical bookstores 0 electronic mail
7" x 10". 0 college bookstores
Hard $23.95 E hobby stores
Paper $16.95 E mail
Book No. 2616 0 art materials stores
Comments
Send $1 for the new TAB Catalog describing over 1300 titles currently in print and receive
a coupon worth S1 off on your next purchase from TAB.
to credit approval Orders outside U S must
(In PA, NY, and ME add applicable sales tax Orders subject
be prepaid with international money orders in U.S dollars.)
'Prices subject to change without notice.
ON IIMM I= MINI NMI MI INN NIB Name
To purchase these or any other books from TAB, visit your local bookstore, Address
return this coupon, or call toll -free 1-800-822-8158 (In PA and AK call
1-717-794-2191). City
Hard or Title Quantity Price
Product No
Payer
State/Zip

TAB BOOKS Inc.

:= Check or money order enclosed made payable to Subtotal

TAB BOOKS Inc.
Postage/
Charge my =2 VISA I: MasterCard :: American Express Handling 52 50
11500 oulsas
Acct. No. Exp us* ssd Casale
in PA NO and ME
Signature add appl.cable
Sales lax
Please P,lni
Name TOTAL

Company
Address Mail coupon to
TAB BOOKS Inc.
City Blue Ridge Summit
State Zip PA 17294-0840 BC
 Logic Made Easy
R. H. Warring

A complete guide to understanding all forms of logic-

from puzzles and problems to algorithms and Boolean Algebra!
Looking for an introductory guide to logic? One that covers the
logic behind puzzles and problems . circuits and binary relation-
.

ships . and the logic of philosophers? Here's an absorbing

. .

sourcebook that will answer all your questions on this fascinating

subject. Author R.H. Warring takes you along the historical path of
considering what logic is all about and how the thought processes
engendered became the building blocks of the new electronic
revolution.
Whether you're an engineer, mathematician, student, or philos-
opher, Logic Made Easy is a valuable tool in solving intellectual or
mathematical and computer -level problems. And it's so easy to un-
derstand that you can explore all forms of logic from Venn . . .

diagrams to Karnaugh Maps.

The early chapters describe types of logic and their use. You'll
explore deductive and inductive logic, logic in solving aptitude tests
and see how to master the use of simple logic diagrams.
And if you're primarily interested in the use of logic as a
mathematical or modern design tool, Warring provides in-depth
coverage of the complex circuits used in microprocessors and com-
puters, complicated control circuits, and computer -aided design.
Seemingly difficult concepts like algorithms, block logic and truth
tables, Boolean Algebra, and even Karnaugh Maps become
crystal-clear.
Drawing on examples and illustrations throughout, Warring fills
a real need by providing a single sourcebook that covers all types
of logic in easy -to -understand English. So whether you're primarily
interested in solving puzzles and mind -teasers understanding . . .

today's modern technology ... or you want insight into the "formal"
logic practiced for centuries by philosophers, here's a guide that
makes it easy to understand and apply.

0885 8 95

32968 LOGIC MADE EitSI t7:;', o! fg-00-00 I ,

035642 891062: C

4 04
.41PRING
LIST PRICE
$8.95
TAB EUERYOAY PRICE
$8
r4orCE
TAB BOOKIfq1u.gs 111111 11
uu
618538
Blue Ridge Summit, PA 17294-0850