-
Notifications
You must be signed in to change notification settings - Fork 0
/
Lecture notes 1 Rings.tm
2970 lines (2505 loc) · 147 KB
/
Lecture notes 1 Rings.tm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<TeXmacs|2.1.2>
<style|<tuple|article|number-long-article>>
<\body>
<\hide-preamble>
<new-remark|observation|Observation>
<assign|remark-name|<\macro|name>
<em|<arg|name>>
</macro>>
<assign|example-text|<macro|<inactive|<localize|Example>>>>
<assign|example|<\macro|body>
<surround|<compound|next-example>||<compound|render-theorem|<compound|example-numbered|<compound|example-text>|<compound|the-example>>|<arg|body>>>
</macro>>
<assign|render-theorem|<\macro|which|body>
<render-enunciation|<theorem-name|<arg|which><theorem-sep>>|<arg|body>>
</macro>>
</hide-preamble>
<section|Rings>
A set <math|R> containing two (possibly equal) elements 0 and 1, and
supporting two binary operations <math|+> and <math|\<times\>> is a
<strong|ring> if
<\itemize-dot>
<item><math|R> equipped with <math|+> is a commutative group;
<item><math|\<times\>> is an associative binary operation on <math|R>
with identity 1;
<item><math|\<times\>> is distributive over <math|+>.
</itemize-dot>
Occasionally we shall have multiple rings and it will be instructive to
clarify which particular ring we are referring to. We shall do this with
subscripts writing, for example, <math|+<rsub|R>> or <math|1<rsub|R>>
instead of <math|+> and 1 above. Identity of associative binary operations
is unique when it exists.
The operation <math|+> is the <strong|addition> of the ring, 0 is the
<strong|zero> of the ring, and the set <math|R> with the operation <math|+>
is the <strong|additive group> of the ring. For each <math|x\<in\>R> we
write <math|-x> for the unique inverse of <math|x> w.r.t. addition, and the
map <math|R\<rightarrow\>R;x\<mapsto\>-x> is the <strong|negation> of the
ring; we write <math|x-y> for <math|x+<around*|(|-y|)>>.
We call a map <math|\<phi\>:R\<rightarrow\>S> between rings
<strong|additive> if it is a homomorphism of the additive groups.
<\observation>
Identities are self-inverse so \<minus\>0 = 0; double inversion is the
identity map i.e. <math|-<around*|(|-x|)>=x> for all <math|x\<in\>R>; and
inversion is a homomorphism of the additive group since a group operation
is commutative (if and) only if inversion is a homomorphism of the group.
Group homomorphisms map identities to identities and inverses to
inverses, so if <math|\<phi\>:R\<rightarrow\>S> is additive then
<math|\<phi\><around*|(|0<rsub|R>|)>=0<rsub|S>> and
<math|\<phi\><around*|(|-x|)>=-\<phi\><around*|(|x|)>> for all
<math|x\<in\>R>.
</observation>
The operation <math|\<times\>> is the <strong|multiplication> of the ring,
and we write <math|x y> in place of <math|x\<times\>y>, and in the absence
of parentheses multiplication precedes addition in the usual way. We say
<math|R> is a <strong|commutative> ring if the multiplication is
commutative.
<\remark>
The modern notion of commutative ring can be traced back to Emmy Noether
[Noe21, \<#00A7\>1] (translated into English in [Ber14]), though unlike
us her definition does not assume the multiplication has an identity;
Poonen [Poo19] defends our position.
</remark>
We call a map <math|\<phi\>:R\<rightarrow\>S> between rings
<strong|multiplicative> if <math|\<phi\><around*|(|x
y|)>=\<phi\><around*|(|x|)>\<phi\><around*|(|y|)>> for all
<math|x,y\<in\>R>, and we call it a <strong|ring homomorphism> if
<math|\<phi\>> is additive, multiplicative, and
<math|\<phi\><around*|(|1<rsub|R>|)>=1<rsub|S>>.
<\observation>
The composition of additive (resp. multiplicative) maps is additive
(resp. multiplicative), and hence the composition of ring homomorphisms
is a ring homomorphism.
</observation>
<\definition>
For a set <math|A\<subset\>X> and a function <math|f:X\<rightarrow\>Y> we
write <math|f<around*|(|A|)>\<assign\><around*|{|f<around*|(|x|)>:x\<in\>A|}>>.
</definition>
For sets <math|A\<subset\>X,B\<subset\>Y>, and a function
<math|X\<times\>Y\<rightarrow\>Z> denoted by infixing the symbol
<math|\<ast\>> between the two arguments, we write
<math|A\<ast\>B\<assign\><around*|{|a\<ast\>b:a\<in\>A,b\<in\>B|}>>; and
denoted by juxtaposing the two arguments, we write <math|A
B\<assign\><around*|{|a b:a\<in\>A,b\<in\>B|}>>.
For <math|x\<in\>X> and <math|y\<in\>Y>, in the case of infix notation we
put <math|x\<ast\>A\<assign\><around*|{|x|}>\<ast\>A> and
<math|A\<ast\>y\<assign\>A\<ast\><around*|{|y|}>>; and in the case of
juxtaposition we put <math|x A\<assign\><around*|{|x|}>A> and <math|A
y\<assign\>A<around*|{|y|}>>.
<subsection|Units and the trivial ring>
An element <math|x\<in\>R> is a <strong|unit> if it is invertible w.r.t.
multiplication <em|i.e.> if there is some <math|y\<in\>R> such that <math|x
y=y x=1>. We write<\footnote>
Some authors (e.g. [Lan02, p84] and [Lam07, xiv]) write
<math|R<rsup|\<ast\>>> for the group of units of <math|R>.
</footnote> <math|U<around*|(|R|)>> for the set of units of <math|R>, and
<math|R<rsup|\<ast\>>> for the set of non-zero elements of <math|R>.
Inverses w.r.t. associative binary operations are unique when they exist,
so for <math|x\<in\>U<around*|(|R|)>> we can unambiguously write
<math|x<rsup|-1>> for the inverse of <math|x>.
<\observation>
Given a ring <math|R>, if <math|x> and <math|y> are units then
<math|<around*|(|y<rsup|-1>x<rsup|-1>|)><around*|(|x y|)>=1=<around*|(|x
y|)><around*|(|y<rsup|-1>x<rsup|-1>|)>> and so <math|x y> is a unit, and
the multiplication on <math|R> restricts to a well-defined binary
operation on <math|U<around*|(|R|)>> making the latter into a group. This
group is called the <strong|group of units> and it has identity
<math|1<rsub|R>> and the inverse of <math|x> in the group
<math|U<around*|(|R|)>> is <math|x<rsup|-1>>, that is the inverse of
<math|x> in the ring <math|R>.
</observation>
<math|1<rsup|-1>=1> since identities are self-inverse; double inversion is
the identity map i.e. <math|<around*|(|x<rsup|-1>|)><rsup|-1>=x> for all
<math|x\<in\>U<around*|(|R|)>>; <math|<around*|(|x
y|)><rsup|-1>=y<rsup|-1>x<rsup|-1>> for all
<math|x,y\<in\>U<around*|(|R|)>>; and if <math|\<phi\>:R\<rightarrow\>S> is
a ring homomorphism then <math|U<around*|(|R|)>\<rightarrow\>U<around*|(|S|)>;x\<mapsto\>\<phi\><around*|(|x|)>>
is a well-defined group homomorphism with
<math|\<phi\><around*|(|x<rsup|-1>|)>=\<phi\><around*|(|x|)><rsup|-1>>.
<\remark>
If <math|R> is a finite commutative ring then <math|U<around*|(|R|)>> is
a finite commutative group, but exactly which finite commutative groups
occur as the group of units of a ring is an open problem called Fuchs'
problem [Fuc58, Problem 72, p299].
</remark>
Given <math|y\<in\>R>, the map <math|R\<rightarrow\>R;x\<mapsto\>y x>
(resp. <math|R\<rightarrow\>R;x\<mapsto\>x y>) is called <strong|left>
(resp. <strong|right>) <strong|multiplication> by <math|y>.
<\observation>
The fact that multiplication is distributive over addition in <math|R> is
exactly to say that all the left and right multiplication maps are group
homomorphisms of the additive group of <math|R>. If <math|x> is a unit
then left and right multiplication by <math|x> are surjective.
Group homomorphisms map identities to identities and inverses to
inverses, so <math|x 0=0 x=0> for all <math|x\<in\>R> \U we say
<strong|zero annihilates>; and <math|x <around*|(|-y|)>=-<around*|(|x
y|)>=<around*|(|-x|)> y> for all <math|x,y\<in\>R> \U we say that
<strong|negation distributes>. In particular <math|<around*|(|-1|)> x=-x>
for all <math|x\<in\>R>.
</observation>
<\example>
The set <math|<around*|{|0|}>>, with <math|1=0>, and addition and
multiplication given by <math|0+0=0\<times\>0=0>, is a ring called the
<strong|trivial> or <strong|zero> ring. A ring in which <math|1\<neq\>0>
is called a <strong|non-trivial> ring.
If <math|R> is not non-trivial then it is trivial: Indeed, since
<math|0=1>, for all <math|x\<in\>R> we have <math|x=1*x=0*x=0> since zero
annihilates and so <math|R=<around*|{|0|}>>. There is only one function
into a set of size one, and so the addition and multiplication on
<math|R> are uniquely determined and must be that of the trivial ring.
</example>
<\example>
The <strong|zero map> <math|z<rsub|R>:R\<rightarrow\><around*|{|0|}>;x\<mapsto\>0>
from a ring <math|R> to the trivial ring is a ring homomorphism.
</example>
<subsection|The integers and characteristic>
We write <math|\<bbb-Z\>> for the integers; <math|\<bbb-N\><rsup|\<ast\>>>
for the positive integers, that is {1, 2, 3, <text-dots>}; and
<math|\<bbb-N\><rsub|0>> for the non-negative integers, that is {0, 1, 2,
<text-dots>}.
<\example>
<math|\<bbb-Z\>> with their usual addition, multiplication, zero and 1
form a non-trivial commutative ring for which
<math|U<around*|(|\<bbb-Z\>|)>=<around*|{|-1,1|}>>.
</example>
<\theorem>
<dueto|The One Ring<strong|<strong|<with|font-series|medium|<\footnote>
Following [Tol04, Book I, Chapter 2, p66] one might describe the
integers as the one ring (up to unique isomorphism) ruling (uniquely
embedding in) all others.
</footnote>>>>>Suppose that <math|R> is a ring. Then there is a unique
ring homomorphism <math|\<chi\><rsub|R>:\<bbb-Z\>\<rightarrow\>R>, and we
have
<\equation*>
\<chi\><rsub|R><around*|(|n-m|)>=<wide|1<rsub|R>+\<cdots\>+1<rsub|R>|\<wide-overbrace\>><rsup|n<text|
times>>-<wide|1<rsub|R>+\<cdots\>+1<rsub|R>|\<wide-overbrace\>><rsup|m<text|
times>>
</equation*>
</theorem>
<\remark>
To prove this we need to define <math|\<bbb-N\><rsup|\<ast\>>> and
<math|\<bbb-Z\>> by Peano axiom, which roughly says <math|\<bbb-Z\>> is
the smallest infinity. The proof is a series of inductions.
</remark>
If there is <math|n\<in\>\<bbb-N\><rsup|\<ast\>>> such that
<math|\<chi\><rsub|R><around*|(|n|)>=0<rsub|R>> then there is a smallest
such <math|n> and we call this the <strong|characteristic> of the ring; if
there is no such <math|n> then the characteristic is said to be 0.
<\example>
For <math|N\<in\>\<bbb-N\><rsup|\<ast\>>>, we write
<math|\<bbb-Z\><rsub|N>> for the integers modulo <math|N>. This is a
commutative ring whose zero is <math|0<pmod|N>>, and with multiplicative
identity <math|1<pmod|N>>. If <math|N=1> then <math|0\<equiv\>1<pmod|N>>
and so the ring is trivial; otherwise it is non-trivial.
</example>
The characteristic of <math|\<bbb-Z\><rsub|N>> is <math|N> since
<math|n\<in\>\<bbb-N\><rsup|\<ast\>>> has
<math|\<chi\><rsub|\<bbb-Z\><rsub|N>><around*|(|n|)>=0<rsub|\<bbb-Z\><rsub|N>>>
if and only if <math|n\<equiv\>0<pmod|N>>, so <math|n\<geqslant\>N> and
<math|\<chi\><rsub|\<bbb-Z\><rsub|N>><around*|(|N|)>=0<rsub|\<bbb-Z\><rsub|N>>>.
<subsection|Isomorphisms and subrings>
A <strong|ring isomorphism> is a map <math|\<phi\>:R\<rightarrow\>S> that
is a ring homomorphism with an inverse that is a ring homomorphism.
<\example>
The identity map <math|\<iota\><rsub|R>:R\<rightarrow\>R;x\<mapsto\>x> is
a ring isomorphism.
</example>
A ring <math|S> is a <strong|subring> of a ring <math|R> if the inclusion
map <math|j:S\<rightarrow\>R;s\<mapsto\>s> is a well-defined \U all this
does is ensure that <math|S\<subset\>R> \U ring homomorphism called the
<strong|inclusion homomorphism>; <math|S> is <strong|proper> if
<math|S\<neq\>R>.
<\example>
<math|\<bbb-C\>> with its usual addition, multiplication, zero and 1 is a
non-trivial commutative ring and <math|\<bbb-Z\>> is a subring of
<math|\<bbb-C\>>.
</example>
<\observation>
The 0 and 1 of a subring are the same as for the containing ring and so a
subring of a non-trivial ring is non-trivial, and the characteristic of a
subring is the same as the characteristic of the ring it is contained in.
</observation>
<with|font|Segoe UI Emoji|\<#26A0\>>In particular, the trivial ring is
<em|not> a subring of any non-trivial ring <math|R> despite the fact that
the inclusion map taking 0 to <math|0<rsub|R>> is both additive and
multiplicative. It follows that the requirement that ring homomorphisms
send 1 to 1 cannot be dropped from the definition.
<\proposition>
<dueto|Subring test>Suppose that <math|R> is a ring and
<math|S\<subset\>R> has <math|1\<in\>S> and <math|x-y,x y\<in\>S> for all
<math|x,y\<in\>S>. Then the addition and multiplication on <math|R>
restrict to well-defined operations on <math|S> giving it the structure
of a subring of <math|R>.
</proposition>
<\proof>
First <math|S> is non-empty and <math|x-y\<in\>S> whenever
<math|x,y\<in\>S> so by the subgroup test addition on <math|R> restricts
to a well-defined binary operation on <math|S> giving it the structure of
a commutative group. Since <math|S> is closed under multiplication,
multiplication on <math|R> restricts to a well-defined binary operation
on <math|S> that is <em|a fortiori> associative and distributive, and
since <math|1\<in\>S> and 1 is <em|a fortiori> an identity for <math|S>,
we have that <math|S> with these restricted operations is a ring. The map
<math|S\<rightarrow\>R;s\<mapsto\>s> is then well-defined since <math|S>
is a subset of <math|R>, and a ring homomorphism as required.
</proof>
Given a subset satisfying the hypotheses of the above lemma, we make the
common abuse of calling it a subring on the understanding that we are
referring to the restricted operations described by the lemma.
<\example>
For <math|d\<in\>\<bbb-N\><rsup|\<ast\>>> we write
<math|\<bbb-Z\><around*|[|<sqrt|-d>|]>> for the set
<math|<around*|{|z+w<sqrt|-d>:z,w\<in\>\<bbb-Z\>|}>>, which is a subring
of <math|\<bbb-C\>> by the subring test. <math|\<bbb-Z\><around*|[|i|]>>
\U the case <math|d=1> \U is called the set of <strong|Gaussian
integers>.
</example>
We have <math|U<around*|(|\<bbb-Z\><around*|[|i|]>|)>=<around*|{|1,-1,i,-i|}>>:
Certainly all the elements of <math|<around*|{|1,-1,i,-i|}>> are units. In
the other direction, suppose <math|<around*|(|z+w*i|)><around*|(|x+y*i|)>=1>
for some <math|x,y\<in\>\<bbb-Z\>>. Taking absolute values we have
<math|<around*|(|z<rsup|2>+w<rsup|2>|)><around*|(|x<rsup|2>+y<rsup|2>|)>=1>,
so <math|z<rsup|2>+w<rsup|2>=1>, and hence
<math|<around*|(|z,w|)>\<in\><around*|{|<around*|(|1,0|)>,<around*|(|-1,0|)>,<around*|(|0,1|)>,<around*|(|0,-1|)>|}>>
as required.
For <math|d\<gtr\>1> we have <math|U<around*|(|\<bbb-Z\><around*|[|<sqrt|-d>|]>|)>=<around*|{|-1,1|}>>
since certainly 1 and \<minus\>1 are units, and if <math|z+w<sqrt|-d>> is a
unit then taking absolute values as above we get <math|x,y\<in\>\<bbb-Z\>>
such that <math|<around*|(|z<rsup|2>+d w<rsup|2>|)><around*|(|x<rsup|2>+d
y<rsup|2>|)>=1>; since <math|d\<gtr\>1> we get <math|w=0> and
<math|z\<in\><around*|{|-1,1|}>> as required.
<\example>
Given a ring <math|R> we write <math|Z<around*|(|R|)>> for the
<strong|centre> of <math|R>, that is the set of <math|x\<in\>R> that
commute with all other elements of <math|R> <em|i.e.> such that
<math|x*y=y*x> for all <math|y\<in\>R>.
The centre is a subring by the subring test: <math|1*x=x=x*1> for all
<math|x\<in\>R>, so <math|1\<in\>Z<around*|(|R|)>>. Secondly, for
<math|x,x<rprime|'>\<in\>Z<around*|(|R|)>>, and <math|y\<in\>R> we have
<math|<around*|(|x-x<rprime|'>|)>y=x*y+<around*|(|-x<rprime|'>|)>y=x*y+x<rprime|'><around*|(|-y|)>=y*x+<around*|(|-y|)>x<rprime|'>=y*x+y<around*|(|-x<rprime|'>|)>=y<around*|(|x-x<rprime|'>|)>>
and <math|<around*|(|x*x<rprime|'>|)>y=x<around*|(|x<rprime|'>*y|)>=x<around*|(|y*x<rprime|'>|)>=<around*|(|x*y|)>x<rprime|'>=<around*|(|y*x|)>x<rprime|'>=y<around*|(|x*x<rprime|'>|)>>.
</example>
<\example>
The ring of integers has no proper subrings, since any such subring must
contain 1 and so by induction <math|\<bbb-N\><rsup|\<ast\>>> and hence
<math|\<bbb-N\><rsup|\<ast\>>-\<bbb-N\><rsup|\<ast\>>=\<bbb-Z\>>.
<with|font|Segoe UI Emoji|\<#26A0\>>The set
<math|\<bbb-N\><rsup|\<ast\>>> contains 1 and if
<math|x,y\<in\>\<bbb-N\><rsup|\<ast\>>> then
<math|x+y,x*y\<in\>\<bbb-N\><rsup|\<ast\>>>, but
<math|\<bbb-N\><rsup|\<ast\>>> is <em|not> a subring of <math|\<bbb-Z\>>
because it does not contain 0. It follows that <math|x-y> may not be
replaced by <math|x+y> in the hypotheses of the subring test.
</example>
<\observation>
For <math|\<phi\>:R\<rightarrow\>S> a ring homomorphism, <math|Im
\<phi\>> is a subring of <math|S> by the subring test:
<math|1<rsub|S>=\<phi\><around*|(|1<rsub|R>|)>\<in\>Im \<phi\>>; and if
<math|x,y\<in\>Im \<phi\>> then there are <math|z,w\<in\>R> such that
<math|x=\<phi\><around*|(|z|)>> and <math|y=\<phi\><around*|(|w|)>> so
<math|x*y=\<phi\><around*|(|z*w|)>\<in\>Im \<phi\>> and
<math|x-y=\<phi\><around*|(|x|)>-\<phi\><around*|(|y|)>=\<phi\><around*|(|x-y|)>\<in\>Im
\<phi\>>.
</observation>
<subsection|Fields>
We say that a commutative ring <math|R> is a <strong|field> if
<math|U<around*|(|R|)>=R<rsup|\<ast\>>>. A subring that is also a field is
called a <strong|subfield>. Throughout these notes <math|\<bbb-F\>> always
denotes a field.
<\example>
The complex numbers <math|\<bbb-C\>> are a field with <math|\<bbb-R\>> as
a subfield.
</example>
<\proposition>
Suppose that <math|\<phi\>:\<bbb-F\>\<rightarrow\>R> is a ring
homomorphism and <math|R> is non-trivial. Then <math|\<phi\>> is an
injection and <math|Im \<phi\>> is a subfield of <math|R>.
</proposition>
<\proof>
If <math|\<phi\><around*|(|x|)>=\<phi\><around*|(|y|)>> and
<math|x\<neq\>y> then <math|x-y\<in\>\<bbb-F\><rsup|\<ast\>>> and so
there is <math|u> such that <math|<around*|(|x-y|)>u=1> whence
<math|0=0*\<phi\><around*|(|u|)>=<around*|(|\<phi\><around*|(|x|)>-\<phi\><around*|(|y|)>|)>\<phi\><around*|(|u|)>=\<phi\><around*|(|<around*|(|x-y|)>u|)>=\<phi\><around*|(|1|)>=1>,
which contradicts the non-triviality of <math|R>.
<math|Im \<phi\>> is a subring of <math|R> which is non-trivial since
<math|R> is non-trivial, and it is commutative since
<math|\<phi\><around*|(|x|)>\<phi\><around*|(|y|)>=\<phi\><around*|(|x
y|)>=\<phi\><around*|(|y x|)>=\<phi\><around*|(|y|)>\<phi\><around*|(|x|)>>.
If <math|\<phi\><around*|(|y|)>\<neq\>0> then since
<math|\<phi\><around*|(|0|)>=0> and <math|\<phi\>> is an injection,
<math|y\<neq\>0> and so <math|y<rsup|-1>> exists and
<math|\<phi\><around*|(|y|)>\<phi\><around*|(|y<rsup|-1>|)>=\<phi\><around*|(|1|)>=1>,
whence <math|\<phi\><around*|(|y|)>> is a unit and <math|Im \<phi\>> is a
subfield.
</proof>
<\warning*>
A subfield doesn't have to be contained in a field.
</warning*>
<\proposition>
Suppose that <math|\<phi\>:\<bbb-F\>\<rightarrow\>R> is a ring
homomorphism. Then the map <math|\<bbb-F\>\<times\>R\<rightarrow\>R;<around*|(|\<lambda\>,r|)>\<mapsto\>\<lambda\>.r\<assign\>\<phi\><around*|(|\<lambda\>|)>r>
is a scalar multiplication of the field <math|\<bbb-F\>> on the additive
group of <math|R> giving an <math|\<bbb-F\>>-vector space such that the
right multiplication maps on <math|R> are linear, and if <math|\<phi\>>
maps <math|\<bbb-F\>> into the centre of <math|R> then so are the left
multiplication maps. In particular if <math|R> is commutative then the
left and right multiplication maps are linear.
Conversely, if <math|R> is a ring which is also an
<math|\<bbb-F\>>-vector space in such a way that all the right
multiplication maps are linear then the map
<math|\<bbb-F\>\<rightarrow\>R;\<lambda\>\<mapsto\>\<lambda\>.1<rsub|R>>
is a ring homomorphism and if all the left multiplication maps are also
linear then its image is in the centre of <math|R>.
</proposition>
<\proof>
The additive group of <math|R> is a commutative group by definition. We
also have <math|<around*|(|\<lambda\>\<mu\>|)>.v=\<phi\><around*|(|\<lambda\>*\<mu\>|)>v=<around*|(|\<phi\><around*|(|\<lambda\>|)>\<phi\><around*|(|\<mu\>|)>|)>v=\<phi\><around*|(|\<lambda\>|)><around*|(|\<phi\><around*|(|\<mu\>|)>v|)>=\<lambda\>.<around*|(|\<mu\>.v|)>;1<rsub|\<bbb-F\>>.v=\<phi\><around*|(|1<rsub|\<bbb-F\>>|)>v=1<rsub|R>
v=v;<around*|(|\<lambda\>+\<mu\>|)>.v=\<phi\><around*|(|\<lambda\>+\<mu\>|)>v=<around*|(|\<phi\><around*|(|\<lambda\>|)>+\<phi\><around*|(|\<mu\>|)>|)>v=\<phi\><around*|(|\<lambda\>|)>v+\<phi\><around*|(|\<mu\>|)>v=\<lambda\>.v+\<mu\>.v>;
and <math|\<lambda\>.<around*|(|v+w|)>=\<phi\><around*|(|\<lambda\>|)><around*|(|v+w|)>=\<phi\><around*|(|\<lambda\>|)>v+\<phi\><around*|(|\<lambda\>|)>w=\<lambda\>.v+\<lambda\>.w>.
It follows <math|R> is an <math|\<bbb-F\>>-vector space as claimed.
Right multiplication by <math|r> is linear since it is a group
homomorphism and <math|\<lambda\>.<around*|(|v*r|)>=\<phi\><around*|(|\<lambda\>|)><around*|(|v*r|)>=<around*|(|\<phi\><around*|(|\<lambda\>|)>v|)>r=<around*|(|\<lambda\>.v|)>r>.
Finally, left multiplication by <math|r> is a group homomorphism, and if
it commutes with all elements of the image of <math|\<phi\>> (which is
certainly true if <math|\<phi\>> maps to the centre of <math|R>), then
<math|\<lambda\>.<around*|(|r*v|)>=\<phi\><around*|(|\<lambda\>|)><around*|(|r
v|)>=<around*|(|\<phi\><around*|(|\<lambda\>|)>r|)>v=<around*|(|r*\<phi\><around*|(|\<lambda\>|)>|)>v=r<around*|(|\<phi\><around*|(|\<lambda\>|)>v|)>=r<around*|(|\<lambda\>.v|)>>,
and so left multiplication by <math|r> is linear.
Conversely, write <math|\<phi\>> for the given map then
<math|\<phi\><around*|(|1<rsub|\<bbb-F\>>|)>=1<rsub|\<bbb-F\>>.1<rsub|R>=1<rsub|R>;\<phi\><around*|(|x+y|)>=<around*|(|x+y|)>.1<rsub|R>=x.1<rsub|R>+y.1<rsub|R>=\<phi\><around*|(|x|)>+\<phi\><around*|(|y|)>>;
and <math|\<phi\><around*|(|x*y|)>=<around*|(|x*y|)>.1<rsub|R>=x.<around*|(|y.1<rsub|R>|)>=x.\<phi\><around*|(|y|)>=x.<around*|(|1<rsub|R>*\<phi\><around*|(|y|)>|)>=<around*|(|x.1<rsub|R>|)>\<phi\><around*|(|y|)>=\<phi\><around*|(|x|)>\<phi\><around*|(|y|)>>
since the map <math|R\<rightarrow\>R;z\<mapsto\>z*\<phi\><around*|(|y|)>>
is linear. It follows that <math|\<phi\>> is a ring homomorphism as
claimed. If all left multiplication maps are linear then for
<math|r\<in\>R> we have <math|r*\<phi\><around*|(|\<lambda\>|)>=r<around*|(|\<lambda\>.1<rsub|R>|)>=\<lambda\>.<around*|(|r*1<rsub|R>|)>=\<lambda\>.<around*|(|1<rsub|R>*r|)>=<around*|(|\<lambda\>.1<rsub|R>|)>r=\<phi\><around*|(|\<lambda\>|)>r>
and so <math|\<phi\><around*|(|\<lambda\>|)>\<in\>Z<around*|(|R|)>>.
</proof>
We call the vector space structure of the proposition the
<strong|<math|\<bbb-F\>>-(vector) space structure on <math|R> induced by
<math|\<phi\>>>.
<\example>
The inclusion map <math|\<bbb-R\>\<rightarrow\>\<bbb-C\>> induces the
usual <math|\<bbb-R\>>-vector space structure on the additive group of
<math|\<bbb-C\>>. <math|<around*|{|1,i|}>> is a basis for this vector
space, which is another way of saying that every complex number can be
written uniquely in the form <math|a+b i> for reals <math|a> and
<math|b>.
</example>
<\example>
Complex conjugation, <math|\<bbb-C\>\<rightarrow\>\<bbb-C\>;z\<mapsto\><wide|z|\<bar\>>>
is a ring homomorphism that is different from the identity so we have two
different <math|\<bbb-C\>>-vector space structures on the additive group
of <math|\<bbb-C\>>: one has scalar multiplication
<math|\<lambda\>.z\<assign\>\<lambda\>*z> and the other
<math|\<lambda\>.z\<assign\><wide|\<lambda\>|\<bar\>>*z> for all
<math|\<lambda\>,z\<in\>\<bbb-C\>>. <hlink|smallest rigid extension field
of the complex numbers|https://mathoverflow.net/questions/61058/>
</example>
<subsection|Zero divisors and integral domains>
In a ring <math|R> we call <math|y\<in\>R> a <strong|left> (resp.
<strong|right>) <strong|zero-divisor> if the left (resp. right)
multiplication-by-<math|y> map has a non-trivial kernel <em|i.e.> if there
is some <math|x\<neq\>0> such that <math|y*x=0> (resp. <math|x*y=0>). A
non-trivial commutative ring <math|R> is an <strong|integral domain> if it
has no non-zero zero-divisors.
<\example>
<math|\<bbb-Z\>> is an integral domain \U it is our prototypical example.
</example>
<\observation>
If <math|x\<in\>U<around*|(|R|)>> then <math|x> is not a left (resp.
right) zero-divisor since if <math|x*y=0> (resp. <math|y*x=0>) then
<math|0=x<rsup|-1>*0=x<rsup|-1><around*|(|x*y|)>=1*y=y> (resp.
<math|0=0*x<rsup|-1>=<around*|(|y*x|)>x<rsup|-1>=y*1=y>).
</observation>
<\example>
Every field <math|\<bbb-F\>> is an integral domain since it is certainly
a non-trivial commutative ring and every non-zero element is a unit and
so not a zero-divisor.
</example>
<\example>
<dueto|Example 1.13, contd.>By Bezout's Lemma if
<math|gcd<around*|(|a,N|)>=1> then there are
<math|\<alpha\>,\<beta\>\<in\>\<bbb-Z\>> such that
<math|\<alpha\>*a+\<beta\>*N=1> and so <math|\<alpha\>*a\<equiv\>1<pmod|
N>>. Since <math|\<bbb-Z\><rsub|N>> is commutative it follows that
<math|a*\<alpha\>\<equiv\>1<pmod|N>> and so <math|a> is a unit in
<math|\<bbb-Z\><rsub|N>>. On the other hand, if
<math|gcd<around*|(|a,N|)>\<gtr\>1> then
<math|a\<times\><frac|N|gcd<around*|(|a,N|)>>\<equiv\>0<around*|(|mod
N|)>> and <math|<frac|N|gcd<around*|(|a,N|)>>\<nequiv\>0<around*|(|mod
N|)>>, so <math|a> is a zero-divisor and hence not a unit. It follows
that <math|U<around*|(|\<bbb-Z\><rsub|N>|)>=<around*|{|a<around*|(|mod
N|)>:gcd<around*|(|a,N|)>=1|}>>.
If <math|p\<gtr\>1> is prime then for all <math|a>, either
<math|p\<divides\>a> or <math|gcd<around*|(|a,p|)>=1>. Hence
<math|U<around*|(|\<bbb-Z\><rsub|p>|)>=\<bbb-Z\><rsup|\<ast\>><rsub|p>>
and so <math|\<bbb-Z\><rsub|p>> is a field; we denote it
<math|\<bbb-F\><rsub|p>> to emphasise this fact.
If <math|N\<gtr\>1> is composite, say <math|N=a*b> for <math|a,b\<gtr\>1>
then <math|a*b\<equiv\>0<pmod|N>> but <math|a,b\<nequiv\>0<pmod|N>> and
so <math|\<bbb-Z\><rsub|N>> is not even an integral domain.
If <math|N=1> then <math|\<bbb-Z\><rsub|N>> is trivial, and so not even
non-trivial!
</example>
<\observation>
If <math|R> is an integral domain and <math|S> is a subring of <math|R>
then <math|S> is an integral domain: <math|S> is certainly non-trivial
and commutative since <math|R> is, and for <math|y\<in\>S>, the
multiplication-by-<math|y> map on <math|S> is the restriction of the
multiplication-by-<math|y> map on <math|R>, and so if the kernel of the
latter is trivial then so is the kernel of the former.
</observation>
<\example>
For <math|d\<in\>\<bbb-N\><rsup|\<ast\>>>, the ring
<math|\<bbb-Z\><around*|[|<sqrt|-d>|]>>, and in particular the Gaussian
integers, is a subring of <math|\<bbb-C\>> and so an integral domain.
</example>
<\example>
The algebraic integers, denoted <math|<wide|\<bbb-Z\>|\<wide-bar\>>>, are
the complex numbers <math|\<alpha\>> for which there is
<math|d\<in\>\<bbb-N\><rsup|\<ast\>>> and
<math|a<rsub|d-1>,\<ldots\>,a<rsub|0>\<in\>\<bbb-Z\>> such that
<math|\<alpha\><rsup|d>+a<rsub|d-1>\<alpha\><rsup|d-1>+\<cdots\>+a<rsub|1>\<alpha\>+a<rsub|0>=0>.
We shall make use of arguments from the modules part of the course to
show that <math|<wide|\<bbb-Z\>|\<wide-bar\>>> is a subring of
<math|\<bbb-C\>>, and given this we conclude
<math|<wide|\<bbb-Z\>|\<wide-bar\>>> is an integral domain.
<math|<wide|\<bbb-Z\>|\<wide-bar\>>> is not a field since
<math|<frac|1|2>\<nin\><wide|\<bbb-Z\>|\<wide-bar\>>>, because if it were
then there would be <math|a<rsub|d-1>,\<ldots\>,a<rsub|0>\<in\>\<bbb-Z\>>
such that <math|1+2<around*|(|a<rsub|d-1>+\<cdots\>+a<rsub|0>2<rsup|d-1>|)>=0>;
a contradiction.
</example>
<\proposition>
Suppose that <math|R> is a ring with no non-zero zero divisors that is
also a finite dimensional vector space in such a way that left and right
multiplication maps are linear. Then <math|U<around*|(|R|)>=R<rsup|\<ast\>>>,
and in particular if <math|R> is an integral domain then <math|R> is a
field.
</proposition>
<\proof>
For <math|a\<in\>R> the map <math|R\<rightarrow\>R;x\<mapsto\>x*a> is
linear, and since <math|R> is an integral domain it is injective if
<math|a\<in\>R<rsup|\<ast\>>>. Since <math|R> is finite dimensional the
Rank-Nullity Theorem tells us that the map is surjective, and hence there
is <math|x\<in\>R> such that <math|x a=1>. Similarly there is <math|y>
such that <math|a*y=1>, and finally <math|x=x*1=x<around*|(|a*y|)>=<around*|(|x*a|)>y=1*y=y>
so <math|a\<in\>U<around*|(|R|)>> as required.
</proof>
<subsection|Product of rings>
For rings <math|R<rsub|1>,\<ldots\>,R<rsub|n>> the product group
<math|R<rsub|1>\<times\>\<cdots\>\<times\>R<rsub|n>> of the additive groups
of the rings <math|R<rsub|i>> may be equipped with a binary operation
defined by <math|<around*|(|x*y|)><rsub|i>\<assign\>x<rsub|i>y<rsub|i>> for
<math|1\<leqslant\>i\<leqslant\>n> making it into a ring with identity
<math|1=<around*|(|1<rsub|R<rsub|1>>,\<ldots\>,1<rsub|R<rsub|n>>|)>>. This
ring is called the <strong|direct product> of the <math|R<rsub|i>>s.
<\observation>
The group of units of a product ring is equal to the product group of the
groups of units of the rings <em|i.e.>
<math|U<around*|(|R<rsub|1>\<times\>\<cdots\>\<times\>R<rsub|n>|)>=U<around*|(|R<rsub|1>|)>\<times\>\<cdots\>\<times\>U<around*|(|R<rsub|n>|)>>.
</observation>
<\example>
The maps <math|\<pi\><rsub|i>:R<rsub|1>\<times\>\<cdots\>\<times\>R<rsub|n>\<rightarrow\>R<rsub|i>;x\<mapsto\>x<rsub|i>>
are ring homomorphisms called <strong|projection homomorphisms>.
</example>
<\example>
The map <math|R\<rightarrow\>R<rsup|n>;x\<mapsto\><around*|(|x,\<ldots\>,x|)>>
is a ring homomorphism called the <strong|diagonal homomorphism (into
<math|R<rsup|n>>)>.
The diagonal homomorphism <math|\<bbb-F\>\<rightarrow\>\<bbb-F\><rsup|n>>
induces an <math|\<bbb-F\>>-vector space structure on
<math|\<bbb-F\><rsup|n>> which is the usual <math|\<bbb-F\>>-vector space
structure on <math|\<bbb-F\><rsup|n>> <em|i.e.> having scalar
multiplication <math|\<lambda\>.v=<around*|(|\<lambda\>v<rsub|1>,\<ldots\>,\<lambda\>v<rsub|n>|)>>
for <math|\<lambda\>\<in\>\<bbb-F\>> and <math|v\<in\>\<bbb-F\><rsup|n>>.
<with|font|Segoe UI Emoji|\<#26A0\>>The ring <math|\<bbb-F\><rsup|n>> has
more structure than the vector space <math|\<bbb-F\><rsup|n>> because the
former comes with a multiplication.
</example>
<\example>
For <math|R> a ring, <math|R<rsup|2>> is never an integral domain: if
<math|R> is trivial then <math|1<rsub|R<rsup|2>>=<around*|(|1<rsub|R>,1<rsub|R>|)>=<around*|(|0<rsub|R>,0<rsub|R>|)>=0<rsub|R<rsup|2>>>,
so <math|R<rsup|2>> is trivial. Otherwise
<math|<around*|(|0<rsub|R>,1<rsub|R>|)><around*|(|1<rsub|R>,0<rsub|R>|)>=<around*|(|0<rsub|R>,0<rsub|R>|)>=0<rsub|R<rsup|2>>>,
<math|<around*|(|0<rsub|R>,1<rsub|R>|)>,<around*|(|1<rsub|R>,0<rsub|R>|)>\<in\><around*|(|R<rsup|2>|)><rsup|\<ast\>>>
and so these are non-zero zero-divisors.
</example>
<subsection|Prototypical rings>
Groups of symmetries are the prototypes for abstract groups and rings have
a similar prototype in which the underlying set is replaced by a
commutative group.
<\proposition>
Suppose that <math|M> and <math|N> are commutative groups with binary
operations <math|+<rsub|M>> and <math|+<rsub|N>>, and identities
<math|0<rsub|M>> and <math|0<rsub|N>> respectively. Then <math|Hom(M,N)>,
the set of group homomorphisms <math|M\<rightarrow\>N>, is itself a
commutative group under + defined pointwise on <math|Hom(M,N)> by
<\equation*>
<around|(|\<phi\>+\<psi\>|)><around|(|x|)>\<assign\>\<phi\><around|(|x|)>+<rsub|N>*\<psi\><around|(|x|)>*<text|for
all >x\<in\>M,
</equation*>
with identity <math|z:M\<rightarrow\>N;x\<mapsto\>0<rsub|N>>, and the
inverse of <math|\<phi\>> is the pointwise negation, meaning for all
<math|x\<in\>M>, <math|(\<minus\>\<phi\>)(x)> is the inverse of
<math|\<phi\>(x)> in <math|N>.
</proposition>
<\proof>
Suppose that <math|\<phi\>,\<psi\>\<in\>Hom<around*|(|M,N|)>>. Then for
all <math|x,y\<in\>M> we have
<\align*>
<tformat|<table|<row|<cell|<around|(|\<phi\>+\<psi\>|)>*<around|(|x+<rsub|M>y|)>>|<cell|=\<phi\>*<around|(|x+<rsub|M>y|)>+<rsub|N>\<psi\>*<around|(|x+<rsub|M>y|)>>|<cell|>>|<row|<cell|>|<cell|=<around|(|\<phi\><around|(|x|)>+<rsub|N>\<phi\><around|(|y|)>|)>+<rsub|N><around|(|\<psi\><around|(|x|)>+<rsub|N>\<psi\><around|(|y|)>|)>>|<cell|<small|\<phi\>*<text|and
>\<psi\><text| are group homomorphisms>>>>|<row|<cell|>|<cell|=<around|(|\<phi\><around|(|x|)>+<rsub|N>\<psi\><around|(|x|)>|)>+<rsub|N><around|(|\<phi\><around|(|y|)>+<rsub|N>\<psi\><around|(|y|)>|)>>|<cell|<small|<text|associativity
and commutativity of >+<rsub|N>>>>|<row|<cell|>|<cell|=<around|(|\<phi\>+\<psi\>|)><around|(|x|)>+<rsub|N><around|(|\<phi\>+\<psi\>|)><around|(|y|)>>|<cell|<small|<text|definition
of pointwise addition>>>>>>
</align*>
It follows that <math|\<phi\>+\<psi\>\<in\>Hom<around|(|M,N|)>>.
Pointwise addition is commutative and associative because addition on
<math|N> is commutative and associative. The map <math|z> is a
homomorphism because <math|z<around|(|x|)>+<rsub|N>z<around|(|y|)>=0<rsub|N>+<rsub|N>0<rsub|N>=0<rsub|N>=z*<around|(|x+<rsub|M>y|)>>,
and it is an identity for pointwise addition because <math|0<rsub|N>> is
an identity for <math|N>. Finally, if <math|\<phi\>\<in\>Hom(M,N)> then
<math|\<minus\>\<phi\>\<in\>Hom(M,N)> because it is the composition of
the homomorphism <math|\<phi\>> and negation which is a homomorphism on
<math|N> since <math|+<rsub|N>> is commutative, and this map is an
inverse for <math|\<phi\>(x)> under pointwise addition by design.
</proof>
<\remark>
To show that <math|Hom<around|(|M,N|)>> is closed under pointwise
addition and negation it is essential that <math|N> be commutative.
</remark>
<\proposition>
Suppose that <math|M,N,> and <math|P> are commutative groups, and
<math|+<rsub|N>> and <math|+<rsub|P>> are the group operations on
<math|N> and <math|Hom(M,N)>, and <math|P> and <math|Hom(N,P)>
respectively. If <math|\<phi\>\<in\>Hom(M,N)> and
<math|\<psi\>\<in\>Hom(N,P)>, then <math|\<psi\>\<circ\>\<phi\>\<in\>Hom(M,P)>;
if <math|\<pi\>\<in\>Hom(M,N)> then <math|\<psi\>\<circ\>(\<phi\>+<rsub|N>\<pi\>)=(\<psi\>\<circ\>\<phi\>)+<rsub|P>(\<psi\>\<circ\>\<pi\>)>;
and if <math|\<pi\>\<in\>Hom(N,P)> then
<math|(\<psi\>+<rsub|P>\<pi\>)\<circ\>\<phi\>=(\<psi\>\<circ\>\<phi\>)+<rsub|P>(\<pi\>\<circ\>\<phi\>)>.
</proposition>
<\proof>
The composition of homomorphisms is a homomorphism which says exactly
that if <math|\<phi\>\<in\>Hom(M,N)> and <math|\<psi\>\<in\>Hom(N, P)>,
then <math|\<psi\>\<circ\>\<phi\>\<in\>Hom(M,P)>. Now, if
<math|\<phi\>,\<pi\>\<in\>Hom(M,N)> and <math|\<psi\>\<in\>Hom(N,P)>,
then
<\equation*>
\<psi\>\<circ\>(\<phi\>+<rsub|N>\<pi\>)(x)=\<psi\>(\<phi\>(x)+<rsub|N>\<pi\>(x))=\<psi\>(\<phi\>(x))+<rsub|P>\<psi\>(\<pi\>(x))=((\<psi\>\<circ\>\<phi\>)+<rsub|P>(\<psi\>\<circ\>\<pi\>))(x)
</equation*>
by definition and the fact that <math|\<psi\>> is a homomorphism, and we
have that <math|\<psi\>\<circ\>(\<phi\>+<rsub|N>
\<pi\>)=(\<psi\>\<circ\>\<phi\>)+<rsub|P>(\<psi\>\<circ\>\<pi\>)> as
claimed. On the other hand, if <math|\<phi\>\<in\>Hom(M,N)> and
<math|\<psi\>,\<pi\>\<in\>Hom(N,P)>, then
<\equation*>
(\<psi\>+<rsub|P> \<pi\>)\<circ\>\<phi\>(x)=\<psi\>(\<phi\>(x))+<rsub|P>\<pi\>(\<phi\>(x))=((\<psi\>\<circ\>\<phi\>)+<rsub|P>(\<pi\>\<circ\>\<phi\>))(x)
</equation*>
by definition. The result is proved.
</proof>
<\remark>
For the identity <math|\<psi\>\<circ\>(\<phi\>+<rsub|N>\<pi\>)=(\<psi\>\<circ\>\<phi\>)+<rsub|P>(\<psi\>\<circ\>\<pi\>)>
we used the homomorphism property of <math|\<psi\>>, while the identity
<math|(\<psi\>+<rsub|P>\<pi\>)\<circ\>\<phi\>=(\<psi\>\<circ\>\<phi\>)+<rsub|P>(\<pi\>\<circ\>\<phi\>)>
followed simply from the definition; <em|c.f.> Exercise I.1.
</remark>
<\theorem>
Suppose that <math|M> is a commutative group. Then the set
<math|Hom<around*|(|M,M|)>> equipped with pointwise addition as its
addition and functional composition as its multiplication is a ring whose
zero is the map <math|M\<rightarrow\>M;x\<mapsto\>0<rsub|M>> and whose
multiplicative identity is the map <math|M\<rightarrow\>M;x\<mapsto\>x>.
</theorem>
<\proof>
<math|Hom<around*|(|M,M|)>> is a commutative group with the given
identity under this addition, and by the second part the proposed
multiplication distributes over this addition. It remains to note that
composition of functions is associative so the proposed multiplication is
associative, and the map <math|M\<rightarrow\>M;x\<mapsto\>x> is
certainly a homomorphism and an identity for composition.
</proof>
<subsection|Matrix rings>
Given a ring <math|R>, we write <math|M<rsub|n,m><around*|(|R|)>> for the
set of <math|n\<times\>m> matrices with entries in <math|R>, and
<math|M<rsub|n><around*|(|R|)>\<assign\>M<rsub|n,n><around*|(|R|)>>. For
<math|A,A<rprime|'>\<in\>M<rsub|n,m><around*|(|R|)>> and
<math|B\<in\>M<rsub|m,p>(R)> we define matrices
<math|A+A<rprime|'>\<in\>M<rsub|n,m><around*|(|R|)>> and
<math|A*B\<in\>M<rsub|n,p>(R)> by
<\equation>
(A+A<rprime|'>)<rsub|i,j>\<assign\>A<rsub|i,j>+B<rsub|i,j><text| and
><around*|(|A*B|)><rsub|i,k>\<assign\><big|sum><rsub|j=1><rsup|m>A<rsub|i,j>B<rsub|j,k>
</equation>
We write <math|0<rsub|n\<times\>m>> for the matrix in
<math|M<rsub|n,m><around*|(|R|)>> with <math|0<rsub|R>> in every entry, and
<math|I<rsub|n>> for the <math|n\<times\>n> matrix with <math|1<rsub|R>>s
on the diagonal and <math|0<rsub|R>>s elsewhere.
<\proposition>
<dueto|Algebra of matrices>Suppose that <math|R> is a ring. Then
<math|M<rsub|n,m><around*|(|R|)>> is a commutative group under + with
identity <math|0<rsub|n\<times\>m>> and for which the inverse of
<math|A\<in\>M<rsub|n,m>(R)> is the matrix <math|\<minus\>A> with
<math|(\<minus\>A)<rsub|i,j>=\<minus\>A<rsub|i,j>>. Furthermore, if
<math|><math|A\<in\>M<rsub|n,m><around*|(|R|)>,B,B<rprime|'>\<in\>M<rsub|m,l><around*|(|R|)>>,
and <math|C,C<rprime|'>\<in\>M<rsub|p,n>(R)> then
<math|C(A*B)=(C*A)B,A(B+B<rprime|'>)=(A*B)+(A*B<rprime|'>),(C+C<rprime|'>)A=(C*A)+(C<rprime|'>A),A*I<rsub|m>=A>
and <math|I<rsub|n>A=A>.
</proposition>
<\remark>
We omit the proof. One can proceed directly using a change of variables
and distributivity, or using some of the language of modules.
</remark>
The commutative group <math|M<rsub|n><around*|(|R|)>> with multiplication
<math|M<rsub|n>(R)\<times\>M<rsub|n><around*|(|R|)>\<rightarrow\>M<rsub|n><around*|(|R|)>;(A,
B)\<mapsto\>A*B> is a ring with multiplicative identity <math|I<rsub|n>> as
a result of the algebra of matrices. A <strong|matrix ring> is a subring of
<math|M<rsub|n><around*|(|R|)>>.
<\remark>
For <math|A\<in\>M<rsub|n><around*|(|R|)>> the <strong|determinant> of
<math|A> is defined to be
<\equation*>
det A\<assign\><big|sum><rsub|\<sigma\>\<in\>S<rsub|n>>sign(\<sigma\>)A<rsub|1,\<sigma\>(1)>\<cdots\>A<rsub|n,\<sigma\>(n)>
</equation*>
where <math|S<rsub|n>> is the symmetry group of permutations of
<math|{1,\<ldots\>,n}>, and <math|sign(\<sigma\>)> is the sign of the
permutation <math|\<sigma\>>.
We shall see in the second half of the course that for <math|R>
commutative, <math|A\<in\>U(M<rsub|n><around*|(|R|)>)> if and only if
<math|det A\<in\>U(R)>, generalising what we already know for fields
since <math|det A\<in\>U(\<bbb-F\>)> if and only if <math|det
A\<neq\>0<rsub|\<bbb-F\>>>. <math|U(M<rsub|n><around*|(|\<bbb-F\>|)>)=GL<rsub|n><around*|(|\<bbb-F\>|)>>
is a field. For non-commutative rings Exercise I.5 gives an example to
show that this equivalence can fail.
</remark>
<\example>
For <math|R> non-trivial the ring <math|M<rsub|2>(R)> is not commutative:
<\equation*>
<around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|1>|<cell|1>>|<row|<cell|0>|<cell|1>>>>>|)><around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|1>|<cell|0>>|<row|<cell|1>|<cell|1>>>>>|)>=<around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|2>|<cell|1>>|<row|<cell|1>|<cell|1>>>>>|)>\<neq\><around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|1>|<cell|1>>|<row|<cell|1>|<cell|2>>>>>|)>=<around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|1>|<cell|0>>|<row|<cell|1>|<cell|1>>>>>|)><around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|1>|<cell|1>>|<row|<cell|0>|<cell|1>>>>>|)>
</equation*>
</example>
<\example>
Given a ring <math|R>, the map
<\equation*>
\<#2206\>:R\<rightarrow\>M<rsub|n><around*|(|R|)>;\<lambda\>\<mapsto\><around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|3|3|cell-halign|c>|<cwith|1|-1|4|4|cell-halign|c>|<cwith|1|-1|4|4|cell-rborder|0ln>|<table|<row|<cell|\<lambda\>>|<cell|0>|<cell|\<cdots\>>|<cell|0>>|<row|<cell|0>|<cell|\<ddots\>>|<cell|\<ddots\>>|<cell|\<vdots\>>>|<row|<cell|\<vdots\>>|<cell|\<ddots\>>|<cell|\<ddots\>>|<cell|0>>|<row|<cell|0>|<cell|\<cdots\>>|<cell|0>|<cell|\<lambda\>>>>>>|)>
</equation*>
is a ring homomorphism called the <strong|diagonal homomorphism (into
<math|M<rsub|n><around*|(|R|)>>)>.
The diagonal homomorphism into <math|M<rsub|n><around*|(|\<bbb-F\>|)>>
induces the usual <math|\<bbb-F\>>-vector space structure on
<math|M<rsub|n><around*|(|\<bbb-F\>|)>> with scalar multiplication
<math|(\<lambda\>.A)<rsub|i,j>=\<lambda\>A<rsub|i,j>>. Writing
<math|E<rsup|<around*|(|i,j|)>>> for the matrix with
<math|E<rsup|<around*|(|i,j|)>><rsub|i,j>=1> and
<math|E<rsup|<around*|(|i,j|)>><rsub|k,l>=0> for
<math|(k,l)\<neq\>(i,j)>, the set <math|{E<rsup|<around*|(|i,j|)>>:1\<leqslant\>i,j\<leqslant\>n}>
is a basis for this vector space.
</example>
<\example>
The <strong|quaternions> are the set
<\equation*>
\<bbb-H\>\<assign\><around*|{|<around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|z>|<cell|w>>|<row|<cell|-<wide|w|\<bar\>>>|<cell|<wide|z|\<bar\>>>>>>>|)>:z,w\<in\>\<bbb-C\>|}>
</equation*>
They form a subring of <math|M<rsub|2><around*|(|\<bbb-C\>|)>> by the
subring test, and in particular <math|\<bbb-H\>> has zero
<math|0<rsub|2\<times\>2>> and multiplicative identity <math|I<rsub|2>>.
Now,
<\equation*>
A\<assign\><matrix|<tformat|<table|<row|<cell|z>|<cell|w>>|<row|<cell|-<wide|w|\<bar\>>>|<cell|<wide|z|\<bar\>>>>>>>\<neq\>0<rsub|2\<times\>2><text|
\ if and only if >det A=<around*|\||z|\|><rsup|2>+<around*|\||w|\|><rsup|2>\<neq\>0
</equation*>
and hence if <math|A\<in\>\<bbb-H\><rsup|\<ast\>>> then the inverse of
<math|A> in <math|M<rsub|2><around*|(|\<bbb-C\>|)>> exists and it is also
in <math|\<bbb-H\>>. Hence <math|A\<in\>U(\<bbb-H\>)> and since
<math|\<bbb-H\>> is non-trivial, <math|U(\<bbb-H\>)=\<bbb-H\><rsup|\<ast\>>>.
<with|font|Segoe UI Emoji|\<#26A0\>><math|\<bbb-H\>> is not a field, or
even an integral domain, since it is not commutative. A not-necessarily
commutative ring in which <math|U<around*|(|R|)>=R<rsup|\<ast\>>> is
called a <strong|division ring> or <strong|skew field>.
Frobenius showed that if a ring is such that every non-zero element is a
unit and it is also a vector space over <math|\<bbb-R\>> in such a way
that left and right multiplication in the ring is linear, then the ring
is isomorphic to either <math|\<bbb-R\>,\<bbb-C\>> or <math|\<bbb-H\>>.
The ring homomorphism
<\equation*>
\<bbb-R\>\<rightarrow\>\<bbb-H\>;\<lambda\>\<mapsto\><matrix|<tformat|<table|<row|<cell|\<lambda\>>|<cell|0>>|<row|<cell|0>|<cell|\<lambda\>>>>>>
</equation*>
has image equal to <hlink|the centre of
<math|\<bbb-H\>>|https://linearalgebras.com/solution-abstract-algebra-exercise-7-1-8.html>,
and so induces a real vector space structure on <math|\<bbb-H\>> in which
left and right multiplication maps are linear. The vector space is
4-dimensional and
<\equation*>
<around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|1>|<cell|0>>|<row|<cell|0>|<cell|1>>>>>|)>,<around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|i>|<cell|0>>|<row|<cell|0>|<cell|-i>>>>>|)>,<around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|0>|<cell|1>>|<row|<cell|-1>|<cell|0>>>>>|)><text|,
and ><around*|(|<tabular*|<tformat|<cwith|1|-1|1|1|cell-halign|c>|<cwith|1|-1|1|1|cell-lborder|0ln>|<cwith|1|-1|2|2|cell-halign|c>|<cwith|1|-1|2|2|cell-rborder|0ln>|<table|<row|<cell|0>|<cell|i>>|<row|<cell|i>|<cell|0>>>>>|)>
</equation*>
form a basis. As element of the group <math|U(\<bbb-H\>)>, these generate
an 8 element subgroup called the <strong|quaternion group> and denoted
<math|Q<rsub|8>>.
There is another natural ring homomorphism: the map
<\equation*>
\<bbb-C\>\<rightarrow\>\<bbb-H\>;\<lambda\>\<mapsto\><matrix|<tformat|<table|<row|<cell|\<lambda\>>|<cell|0>>|<row|<cell|0>|<cell|<wide|\<lambda\>|\<bar\>>>>>>>
</equation*>
which induces a 2-dimensional <math|\<bbb-C\>>-vector space structure on
<math|\<bbb-H\>> in which right multiplication maps are linear, but left
multiplication maps are not (in general).
In fact there is no <math|\<bbb-C\>>-vector space structure on
<math|\<bbb-H\>> such that all left and right multiplication maps are
linear: If there were it would give rise to a ring homomorphism
<math|\<bbb-C\>\<rightarrow\>\<bbb-H\>> mapping into the centre of
<math|\<bbb-H\>>. The centre of <math|\<bbb-H\>> is isomorphic to
<math|\<bbb-R\>>, and hence we could have a ring homomorphism
<math|\<bbb-C\>\<rightarrow\>\<bbb-R\>> which we see in Exercise I.3 is
not possible. <with|font|Segoe UI Emoji|\<#26A0\>>In particular,
<math|\<bbb-H\>> is <em|not> a subspace of the usual
<math|\<bbb-C\>>-vector space <math|M<rsub|2>(\<bbb-C\>)> as defined in
Example 1.47 because in that structure left and right multiplication
<em|are> linear, and since <math|\<bbb-H\>> is a subring it were also
subspace they would restrict to be linear on <math|\<bbb-H\>>.
</example>
<subsection|Polynomial rings>
<\proposition>
<dueto|Algebra of polynomials>Suppose that <math|R> is a subring of
<math|S>, <math|\<lambda\>\<in\>S> commutes with all elements of
<math|R>, and <math|a<rsub|0>, a<rsub|1>,\<ldots\>,b<rsub|0>,b<rsub|1>,\<ldots\>\<in\>R>
have <math|a<rsub|i>=0> for all <math|i\<gtr\>n> and <math|b<rsub|j>=0>
for all <math|j\<gtr\>m>. Then
<\equation*>
<around*|(|<big|sum><rsub|i=0><rsup|n>a<rsub|i>*\<lambda\><rsup|i>|)>+<around*|(|<big|sum><rsub|j=0><rsup|m>b<rsub|j>*\<lambda\><rsup|j>|)>=<big|sum><rsub|i=0><rsup|max
<around|{|n,m|}>><around*|(|a<rsub|i>+b<rsub|i>|)>*\<lambda\><rsup|i><space|1em><text|and><space|1em>-<around*|(|<big|sum><rsub|i=0><rsup|n>a<rsub|i>*\<lambda\><rsup|i>|)>=<big|sum><rsub|i=0><rsup|n><around*|(|-a<rsub|i>|)>*\<lambda\><rsup|i>
</equation*>
and
<\equation*>
<around*|(|<big|sum><rsub|i=0><rsup|n>a<rsub|i>*\<lambda\><rsup|i>|)><around*|(|<big|sum><rsub|j=0><rsup|m>b<rsub|j>*\<lambda\><rsup|j>|)>=<big|sum><rsub|k=0><rsup|n+m><around*|(|<big|sum><rsub|j=0><rsup|k>a<rsub|k-j>*b<rsub|j>|)>*\<lambda\><rsup|k>
</equation*>
</proposition>
<\remark>
We omit the proof though it is not difficult: it makes essential use of
distributivity and changes of variables.
</remark>
For a non-trivial ring <math|R> there is a non-trivial ring
<math|R<around*|[|X|]>> called the <strong|polynomial ring over <math|R>
with variable <math|X>> with <math|R> as a subring, and a distinguished
element <math|X\<in\>R<around*|[|X|]>> which commutes with all elements of
<math|R<around*|[|X|]>>, <em|i.e.> <math|p X=X p> for all
<math|p\<in\>R<around*|[|X|]>>, such that
<\equation>
R[X]={a<rsub|0>+a<rsub|1>X+\<cdots\>+a<rsub|n>X<rsup|n>:n\<in\>\<bbb-N\><rsub|0>,a<rsub|0>,\<ldots\>,a<rsub|n>\<in\>R}
</equation>
and
<\equation>
a<rsub|0>+a<rsub|1>X+\<cdots\>+a<rsub|n>X<rsup|n>=0<rsub|R>\<Rightarrow\>a<rsub|0>,\<ldots\>,a<rsub|n>=0<rsub|R>
</equation>
<\remark>
We omit the proof that such a ring exists, but the idea is to take the
additive group of functions <math|\<bbb-N\><rsub|0>\<rightarrow\>R> with
a finite number of non-zero entries and group operation coordinate-wise
addition, and identify <math|X<rsup|n>> with the function taking <math|m>
to <math|0<rsub|R>> if <math|m\<neq\>n> and <math|1<rsub|R>> if
<math|m=n>.
</remark>
For more variables we define <math|R[X<rsub|1>,\<ldots\>,X<rsub|n>]\<assign\>R[X<rsub|1>,\<ldots\>,X<rsub|n-1>][X<rsub|n>]>
and call it the <strong|polynomial ring over <math|R> in the variables
<math|X<rsub|1>,\<ldots\>,X<rsub|n>>>.
The algebra of polynomials and (1.3) allows the <strong|equating of
coefficients>, meaning that if <math|a<rsub|0>+a<rsub|1>X+\<cdots\>+a<rsub|n>X<rsup|n>=b<rsub|0>+b<rsub|1>X+\<cdots\>+b<rsub|m>X<rsup|m>>
for <math|a<rsub|0>,a<rsub|1>,\<ldots\>,b<rsub|0>,b<rsub|1>,\<ldots\>\<in\>R>
with <math|a<rsub|i>=0> for <math|i\<gtr\>n> and <math|b<rsub|j>=0> for
<math|j\<gtr\>m>, then <math|a<rsub|i>=b<rsub|i>> for all <math|i>.
If <math|p\<in\>R[X]<rsup|\<ast\>>> then there is a minimal
<math|d\<in\>\<bbb-N\><rsub|0>> and unique elements
<math|a<rsub|0>,a<rsub|1>,\<ldots\>,a<rsub|d>\<in\>R> with
<math|a<rsub|d>\<neq\>0<rsub|R>> such that
<math|p(X)=a<rsub|0>+a<rsub|1>X+\<cdots\>+a<rsub|d>X<rsup|d>>. We call this
minimal <math|d> the <strong|degree> of <math|p> and denote it <math|deg
p>; we call <math|a<rsub|i>> the <strong|coefficient> of <math|X<rsup|i>>;
<math|a<rsub|d>> the <strong|lead coefficient> and <math|a<rsub|0>> the
<strong|constant coefficient>.
A polynomial is <strong|monic> if its lead coefficient is 1, and the
<strong|constant polynomials> are those for which the constant coefficient
is the only coefficient that may be non-zero.
<\example>
The inclusion homomorphism <math|\<bbb-F\>\<rightarrow\>\<bbb-F\>[X]>
induces an <math|\<bbb-F\>>-vector space structure on <math|\<bbb-F\>[X]>
in such a way that all multiplication maps are linear. In this space,
(1.2) says exactly that <math|{1,X,X<rsup|2>,\<ldots\>}> is a spanning
set, while (1.3) tells us it is linearly independent.
</example>
<\proposition>
Suppose that <math|\<phi\>:R\<rightarrow\>S> is a ring homomorphism from
a non-trivial ring, and <math|\<lambda\>\<in\>S> commutes with all
elements of the image of <math|\<phi\>>. Then there is a unique ring
homomorphism <math|R<around*|[|X|]>\<rightarrow\>S> extending
<math|\<phi\>> and mapping <math|X> to <math|\<lambda\>>, and we have
<\equation*>
a<rsub|0>+a<rsub|1>X+\<cdots\>+a<rsub|d>X<rsup|d>\<mapsto\>\<phi\>(a<rsub|0>)+\<phi\>(a<rsub|1>)\<lambda\>+\<cdots\>+\<phi\>(a<rsub|d>)\<lambda\><rsup|d>.
</equation*>