Carl Faith, Professor Emeritus, Mathematics, Rutgers University


ERRATA AND ADDENDA CUMULATIVE UPDATE OF APRIL 2000

CARL FAITH  "Rings and Things and a Fine Array of Twentieth Century Associative Algebra,"
Surveys of the A.M.S., vol 65, Providence, 1999.

Google Book Search

 (Note the publishing year '99, although printed in '98. Also note, I did not write the advertisement blurbs. I wanted to have the work of many other mathematicians cited besides the very few that AMS listed.) 

PLEASE SEND ANY CORRECTIONS / SUGGESTIONS TO: CarlFaith@aol.com   

Note: The date of the change is December '98 except as noted in parentheses (month/year).  Please share this with other colleagues in your department by xeroxing copies for them or have them email us for copies sent electronically. A tentative date for a corrected reprint  is 2002. 

REMARK:
Most of the Errata falls into the "annoying but obvious" category, except for the correction 157 / -23, Theorem 9.6 on p. 164, correction 190 / 20, and the definitions on 223 /-9,-10, 238 (preceding 16.33) and 254/14.
 

Page / Line Date Is / Ought 
vii / 18 (1/00) Replace the epitaph in the second dedication by:
So mayest thou, till suddenly, like a ripe fruit drop in thy mother's lap.  - from Paradise Lost by John Milton
xxiv / 5      /
       / 6   + / =
xxvii /-15 (4/00)      (6.99) / (6.9) 
xxviii / 4  (4/00)  R is locally / R is Goldie and locally 
xxx / 7 (4/00) abetted, / abetted 
xxxvi / 15 (12/99) exists / exist 
xxxvii / 11 (12/99) transcendence / transcendence degree 
xxviii / -2 (2/99) v / xxv 
xv /-23 (1/99) Indent "Modules" 
xvii /-18 (1/99) Polynomals / Polynomial 
1 / 10 (12/99) I. / I. Rank need not be unique. See Cohn [77], pp 103-4. 
1 / 15 (1/99) In this case, / In case [S] = M then,
1 / 17 (12/99) /
1 / 18   [76] / [72a] ( [72a] was actually published in '73) 
2 / 9   Index 2. / index 2. (5) The characteristic char R of a ring R is n if there exists a least integer n > 0 such  that n . 1 (= n sums of 1) = 0. Then na = 0 VaeR. ... If no such n exists, then we say char R = 0. 
4 /-10 (1/99) / every  
6 /-15 (2/99)  that / that
6 / (2/99) Add Footnote 2: The term RD (= relatively divisible) was used by Warfield [69a]. See Fuchs, Salce,    and Zanardo [99] for a historical sketch. Cf. 6.46A.
9 / 20 (2/99)  for a prime p. / for a prime p. Then f has characteristic 0 in the first instance, and p in the second. Also  see Kleiner [99] for a sketch of the development of the axioms of fields, including the contributions of Weber, Dedekind, Hensel, Steinitz, Artin and Schreier, among others. 
9 /-3; 10 / 5 (12/99) algebraicaly / algebraically 
15 / 9 (2/99)  ideal. / ideal. Conversely, every simple right R-module V is cyclic, in fact, V = vR for any nonzero  element v in V. Moreover V is isomophic to R/I, where I is the annihilator of v in R, hence a maximal right ideal. 
16 / 12 (12/99)  [91] / [95] 
17 /-11  (3/99)  sup.2.17s8./ sup. 2.17A and 2.17F (pp. 31 and 33). 
18 / 1  (1/99)  algebra / central algebra 
21 /-12  (12/99) Add: k[[x]] also denotes the power series. 
24 / 4  (3/99) P.M. Cohn / P.M. Cohn [61]
25 / 18,19  (2/99)  of a subfield / of a proper subfield 
27 /-12  (5/99)  '         
31 / 2    of a set / of subsets of a set 
31 / 5    that M / that a module M 
31 / 20 (12/99)  Delete S
31 / 21  (12/99)  Add: Note that S is nonempty. 
32 / 7 (12/99)  irreducible / irreducible qua submodule, 
33 / 17, 18   Delete the entire offending clause beginning with "although" and ending with "cyclic." 
34 / 13,14 (12/99)  Preceding the paragraph heading "Hilbert Basis Theorem," add: 
2.19C COROLLARY. If R is commutative Noetherian but non-Artinian ring, then there exists an ideal I maximal with respect to the property that R/I is non-Artinian. Futhermore, any such ideal I is a prime ideal and R/K is Artinian for any ideal K properly containing I. Proof Exercise. 
35/-11 (10/00) number field L./number field L. See, e.g., Osofsky [99] for an elementary proof.
35 /-1  (12/99)  Add to the footnote: P.M. Cohn points out that the "usual spelling is Shafarevich." 
36 / 2 (12/99)   iteralia / inter alia 
38 / 10 (12/99)  co-irreducible / irreducible (formerly called co-irreducible) 
39 / 1 (12/99)  [Z-W] / [ Z-S] 
43 / 16 (10/99)  P / B 
52 /  (12/99) In 3.2D: Schöpf / Schopf 
 53 / 4  (12/99)  1-1 / injective 
55 / 1 (4/00)  THEOREM / THEOREM [60] 
55 /-11 (10/99) cogenerator / generator 
56 / 10  (3/99)  Theorem 3.8 / Theorem 3.7C 
 56/ 17  (12/99)  R. / R.} 
57 / 5   

 (,) / (,)

60 / -15    -injective / -injective (Cf.3.7A) 
61 / 12  (4/00)  Goursaud and Valette / Goursaud and Valette [75], who prove that any ring R with a faithful sigma-injective right R-module has acc on direct sums of right ideals. See finite Goldie dimension below. 
61 / -2  (2/99)  cf. Chapter 8 / cf. Chapter 8 and 16.9B. 
62 / 1  (2/99) in a variable x / in any finite or infinite number of variables 
62 / 6,7  (4/00)  To "REMARKS" add: (4) For some pathology on the Goldie dimension of a sum of two submodules, see Camillo [78b], Camillo and Zelmanowitz [78], and Valle [94]. 
62 / 12  (2/99)  rings. / rings. Shock 72, Theorem 3.6, extended Small's theorem to infinite polynomial rings. 
62 / 15  (4/00)  Add: For theorem 3.13 for rings with involution, see Domokus [94]. 
64 / 1  (12 /99)  An R-module / An injective R-module 
64 / 15    Cf.4.6D / Cf. 7.32s 
64 /  (2/99)  Following 3.15C, add: Remark. Matlis [58] proved that any injective module E over a Noetherian commutative ring R had this structure and that there is a 1-1 correspondence between prime ideals P and indecomposable injectives E(R/P). Cf. Theorem 3.4 above. Also see Goodearl-Warfield [89], Theorem 4.24, p. 79. 
64 / 17    /
64 /  (2/99)  To Theorem 3.16C add: 4. R has Noetherian quotient ring R REMARK 4. is Beck's Corollary 3.10 We come back to these ideas in 16.33. 
64 /  (3/99)  Preceding Theorem 3.17B, add: REMARK. Note that requiring the acc on annhilators in R/I suffices for (3) of 3.17A. Cf. Mori domains, 9.4s, p. 164. 
65 /  (12/99)  In 2. of 3.19A: equivalent / equivalently
66 / (12/99) In 3.24A, delete "f.g" 
66/1 (6/00) Cozzens [70] / Cozzens [70] and Koifman [70]
66/6 (6/00) Resco [87]. / Resco [87].  See Komarnitskii [97] for solution  to a question stated in the title ibid.
67 / 8   ideal / proper ideal 
67 / 17   free. / free. Cf.3.23C 
68 / 1,3  (1/99)  lift/ring should be lift/rad ring 
68 / 2,3  (1/99)  See Jacobson...SBI-rings / Jacobson's SBI-rings in his book [56,64], p.53, are lift/rad rings. 
68 / 12  (12/99)  modules / module 
70 / -18    3.33F R is / 3.33F. R is 
70 /-2  (12/99)  Sätzes / Sätze 
74/-14  (12/99)  add: Graham Higman [56] proved it in greater generality. However, in his review (MR, 92d: 13023) of   Formanek's Survey [90], E. Zel'manov states (without a reference!) that J.Dubnov and V. Ivanov proved the Nagata-Higman Theorem in 1943. 
75 / 5  (2/99)  See / see / 16 (1/99) nilalgebra / nil algebra 
77 / -6    Similar Notation / Similar. Notation 
81 / 10    of M. / of M onto M. 
84 /  (2/99)  Above Kolchin's Theorem, add: Remark. See Shock [72b] for generalizations to modules M of finite Goldie dimension or acc on "rationally closed" submodules (see 12.OB). In the latter case, every nil subring S of A is nilpotent. This generalizes a theorem of Small for Noetherian M. (See op.cit., p.313.) 
     / -18    If g is / If t is 
88 /  (2/99)  Add: 3.81 Theorem (Ibid.) If R is commutative, then prim rad R is essentially nilpotent 
89 / 8  (12/99)  1 + i / VNR(i) 
89 / 14  (12/99)  is / if 
91 /-13,-12  (2/99)  C.V. Jensen / C.U. Jensen 
94/ -1 )2/01) generalizes/ is 4.5A Thereom (Kaplansky [50]) Let R be a VNR ring, and n and interger >0. Then R ~An for an Abelian VNT ring A iff all (left) primitive factor rings have index n.
95/ 8 (2/01) Kaplansky [50] (all R/M have the same index n and / Kaplansky [50] without assuming self-injectivity when all R/M have the same index, and then
95 / 11  (4/00)  Utumi. / Utumi. Also see Carson [76] proves 4.5 for complete VNR maximal left quotient rings of finite 
index. 
95 /-10  (12/99)  Was Sind und Was Sollen Die Zahlen / Was sind und was sollen die Zahlen 
100 /-11,-12  (3/00)  Add preceding 4.16C: REMARK. Sweedler [75] (for commutative algebras) and Lawrence [76] characterize when a tensor product of algebras is local 
101 / 10    4.26C / 4.2C
        / 18  (3/99)  -ring / -ring
102 / 8  (11/99)  13.7 / 13.7. Also see Rutter [71], where this theorem is extended to the endomorphism ring of "PF-modules." 
104 / 14-15    Following line 14, preceding Remark add: 4.32 Theorem (Kitamura[91]). If A is a separable k-algebra whose center is a free k-module, then A is FPF iff k is FPF. Note. This extends results of S. Page for when C=k, and Herbera - Menal [89] for the group algebra A = kG of a finite group of unit order. 
104 / 16    Page[84e] / Page [84] 
106 / 13  (6/99)  113.7A / 13.7A 
108/-9  (4/00) (See 5.4B)./ (see 5/4B). Furthermore:
      5.4(Theorem (Nagata [62], p. 55, 17.7) Let R be a Noetherian commutative semilocal ring with Jacobson radical m, and let (,...) be an ideal whose radical is m. Then the completion R* of R in the m-adic topology is isomorphic to the power series ring A in n variables over R modulo the ideal ( - , ..., - ) of A.
     5.4 Theorem (Chevalley [43]) Let , ... be the maximal ideals of R in Theorme 5.4 .  Then R* is the finite product of the completions of the local rings of R at ,=1,...,n.
109 /-3  (3/00)  VÁMOS [77,79] / VÁMOS [77b,79] 
110 / 4  (3/00)  [76] / [77b] 
111 / 14, 15  (12/99)  Add: 5.14C Remark. According to Kleiner [99], the concept of a valuation of general fields extending Hensel's p-adic valuations is due to Kürshak in 1913, who proved the existence of their completions, and in 1918 Ostrowski determined all valuations of the field Q of rational numbers. 
111 /-15,-16  (12/99) Maclane / Mac Lane 
111 /-2,-3  (4/00) Also see F.K. Schmidt [33] and Schilling [50], Chap. 7, on multiply complete fields. 
112 / -5    requires a theorem / requires ideas of a theorem / -5 See 7.21ff / See 7.21ff and "Letter from Victor Camillo", p. 254. 
118 / 3    Let R be semiperfect / Let R be a semiperfect ring 
118 /-13  (3/99)  Faith-Page [81] / Faith-Page [84] 
119 / 3    Faith-Page [81] / Faith-Page [84]
       / 7  (2/99)  Remark 4.84f / Example 4.24 
120 / 5  (2/99) VD). / VD). Also see 4.24-5. 
121 /-1  (2/99)  Proposition, p. 84) / Proposition, p. 84. In fact, pure-injectivity can be viewed as injectivity in an appropriate category; see Gruson-Jensen [73].) 
123 / 1  (2/99)  50 / some 40 
       /-10  (2/99) Hendriksen / Henriksen 
124 /  (2/99)  Preceding 6.3F, insert: Remarks. (1) Any ring with stable range is Dedekind finite (if xa = 1 then, taking b = 0 in the definition, we see that a is a unit.) (2) The above definition for stable range 1 is right- left symmetric, as T.Y. Lam points out in his Exercises in Classical Ring Theory (Springer Verlag) pp. 15-16, Ex. 1.25 and comment. (I have Lam [99b] to thank for these remarks.) 
124 / -1   Faith [91b] / Faith [91b]. Cf. Theorem 16.31. 
125 / 1  (1/99)  A ring R / A commutative ring R
126/ -7   (IF4), R is then / any VNR ring R is
127/ (11/00) Insert preceeding REMARK 6.14A Theorem. Let R be strongly regular.
(1) (Hirano, Hung and Kim [95]) If R is &o-injective, then R<x> C =R[[x]]) is a duo ring and Bezont.
(2) (Karamzadeh and Koochakkapoor [99].) The converse of (1) holds.
128/ -5   in A. R / in R. R 
128/19,20 (9/00)  Following "REMARK"  add:6.19A Theorem. (Brandal [73], Olberding [99])   Every homomorphic image of the quotient field Q of a domain R is injective iff R is an AMVR Prüfer domain.
REMARK.  See loc.cit. for other results, and see the Math. Rev. (2000f: 13041) of Olberding's paper by M. Fontana for historical background.
129 /-9,-10  (4/00)  Add: Cf. R.E. Johnson [69] for other examples of Mal'cev domains. 
129 /-8  (6/99)  field / sfield 
130  (12/99)  Above 6.31, insert: Definition. A ring R is a right (semi) fir if every (f.g) right ideal is free of unique rank. 
131 / 18  (12/99)  contructions / constructions 
132 / 9  (12/99)  injective / quasi-injective 
132 / 10  (2/99)  (ibid.) / (ibid.) Also, Kawada [57] showed that a right self-pseudo-injective algebra of finite dimension over an algebraically closed field is QF. 
133/ 6  (3/99)  Q(R[X] / Q(R[X]) 
      / 10  (11/99)  intersection, then / intersection. Then
      / 11    Cf. sup. 7.8 / Cf. 3.58-61 and sup. 7.8 
133    preceding "6.39 Theorem", add this paragraph: 

ASSOCIATED PRIMES OF POLYNOMIAL RINGS

For a commutative ring R, an associated prime ideal is a prime ideal P that is the annihilator of some element of R, and Ass R denotes the set of all such P. (See 16.11). By theorem 2.37E, any maximal annihilator ideal T is an associated prime ideal (also see 16.12), and Ass*R denotes the set of all 
such T.

   (9/99)  Replace "6.39 Theorem and Remark" by: 
6.39 Theorem (Brewer-Heinzer [74]). For any commutative ring R, the extension mapping of Ass R to Ass R[X] sending P to P[X] is a bijection. Thus: card Ass R = card Ass R[X] and card Ass*R = Ass* R[X]. Remark. The proof employs the concepts of Lazard [69], and localization theory. The author [00b] has given a direct proof using results of Shock [72b]. 
133 /-12  (12/99)  McCoy[57] / McCoy [57b] 
135 / 7  (2/99)  [98] / [99] 
       / 9  (2/99)  (DVD). / (DVD). This is a corollary to ibid. Theorem 6 that states that "RD-essential" is transitive over a domain R iff R is a DVD. Cf. 6.46A. 
136  (12/99)  Trilfaj / Trlifaj 
136 /-1  (12/99)  1.24B / 1.24B Cf. also Azumaya [96] who points out that if every projective left module is pure-injective, then R must be left perfect. 
137 / 9    direct summands. / direct summands. Cf.8.6s 
139 / 5   restricted Artinian / restricted Artinian. Cf. 2.19B. 
140 /-1  (2/99)  Add: Cf. 13.45(3) 
140  (4/00)  Following Theorem 7.7 add: 
7.7A THEOREM (Cozzens and Faith [75], p. 138) Any right ideal of a simple right hereditary ring can be generated by 2 elements. 7.7B REMARK. See Ibid, p. 33, Lemma 2.24 for a short proof of the first statement of Theorem 7.7. Also see theorem 7.12 below. 
140 /-3  (12/99)  Sandomerski / Sandomierski 
141 / 5,6 (12/99)  Add: 7.8A THEOREM (Jategaonkar [74b] Over a (twosided) fully bounded Noetherian ring R any f.e. module M is Artinian, and if M is also f.g, then M has finite length. Remark. For a commutative Noetherian ring this is a theorem of Matlis [58]. 
141  (4/00)  Add to Theorem 7.12: Any right Goldie simple ring R of gl.dim not exceeding 2 is right Noetherian (Ibid., Theorem 2.25, Proposition 2.38 and Theorem 2.40.) 
145 / -9    Add: By Theorem 3.33A, M is a min module iff semiArtinian. 
146 / 1,2  (2/99)  Shock's Theorem...latter. / Cf. Vic Camillo's Letters, p. 254. 
148 / -9   Cf. 13.62(2) / Cf. Shock's Theorem 7.28 
150 /-7  (10/99) Delete the previous addendum of (3/99).
       / -6,-7  (10/99)  Insert: Theorem. An injective right R-module E is -injective iff E is countermodule of finite length.

Proof. This follows easily from Prop. 3.8(a), and the Teply-Miller Theorem 3.10. Cf. p. 30 of the author's lectures [82a], esp. Corollary 7.5. 
153/14,15 (11/00) Krull [28] - Schmidt [25,26] / Krull [25,26] - Schmidt [28]
156 /13  (1/99)  A ring R...provided that / A ring R is a lift/rad ring provided that
       / 18  (1/99)  R is SBI / lift/rad ring 
157 / 16 (2/99) Harano / Hirano 
       / -18    indempotents / idempotents 
       / -10    isormophic / isomorphic 
       / -23  (1/99)  Replace Example 4 by : Example 4 (Lam). A Dedekind finite exchange ring R need not be semiperfect, e.g., any infinite product R of copies of any field: R is self-injective hence suitable by 4.2A hence an exchange ring by 8.4C. 
158  (2/99)  To Corollary 8.5B, add: (Cf. 8.1s and 8.3.) 
158 /-12  (4/99)  esp. 16.39 / esp. 16.39-16.43 
159 /-6  (4/00)  ring; / ring A; 
159 /-5  (4/00)  R-module / R-module M 
159 /-1  (4/00)  M and N. / M and N. A necessary condition for the truth of the conjecture is that the center of A (= the biendomorphism ring of M) be linearly compact. 
163 / 5  (2/99)  Faith [86] / Faith [86b] 
       / 8  (2/99)  Faith [86] / Faith [89a] 
163 /-3,-4  (4/00)  Add: REMARK. Kerr [79] gives an example of a commutative Goldie Kerr ring whose 2x2 matrix ring is not Goldie. (In this example R has just two nontrivial annihilator ideals and Goldie dimension two! 
Cf. Theorem 9.4 & Corollary 9.5.) 
163 /-3  (3/00)  Roitman [89,I] / Roitman [90] 
164 / 3,4  (3/00)  Insert: REMARK. Roitman applied the coup de grace to Theorem 9.3 in the following: 9.3' THEOREM. Over a countable field k there is an acc algebra that is not Kerr. 
164 /12  (1/99)  [93,94,96] / [91a, 94,96b] 
164 / 16,17 (4/00)  Add: REMARK. Since Q is Artinian in Theorem 9.4 (and the next corollary), then any n x n matrix ring over R satisfies acc , hence is Goldie. 

164 

(1/99)  There is a gap in the proof of Theorem 9.6: In order to invoke Beck's Theorem, Ass R must be finite. The set Ass*R of maximal associated primes is finite in a ring with acc by Theorems 16.25 or 16.31. Is Ass R finite in this case? Since Ass R is finite in a ring of finite Goldie dimension by Corollary 16.18, then "Theorem 9.6" holds assuming, e.g., that R is Goldie. 
Hence replace 9.6 by: 
9.6 Theorem. If R is a commutative Goldie ring, and if the local ring of R at each associated prime ideal P is Noetherian, then R has a flat embedding in a Noetherian ring, hence R is Kerr, and so is any polynomial ring R[X] over R. 
Proof. See Beck's Theorem 3.16C. Also see Theorem 16.33 and Remark 16.34. 
165 /-16  (2/99) [84b,c] / [84b, 86c]
       /-14  (2/99)  [84c] / [86c], p. 40, theorem 2.1 
      /- 13  (2/00) Roitman [94,II] / Roitman [89,II] /-11 (12/99) element of K / element x of K 
166 / 12  (12/99) integral / integrally 
166 /-15  (7/99)  conversely if K / conversely if A 
167 / 6  (2/99)  9.16B / 9.15B 
168 / 16  (6/00)  domains. / domains.  Also see Osmanagic [99] for a general approximation theorem for "non-Marot Krullrings."
170 /13  (10/99)  /
171/-9,-8  (1/99) Put "generalized discrete valuation domain" into bold face type. 
       / -7  (2/99) M / m 
174 / 9  (2/99)  is isomorphic to S / R is isomorphic to S 
174/-5,-4,-3  (1/99)  (1), (2), (3) / (4), (5), (6) 
175 / 9  (12/99)  finite. / finite (and then A is right SI by 11.1.) 
177 / 15  (12/99)  Higman [56] / D.G. Higman [54] 
178 / 3  (3/99)  domain? / domain when K is ? 
179 / 10,11  (2/99)  these rings...defined / these rings R are essential R-submodules 
180 -21 (2/99)  large / essential 
181 / -15    Right singular / Right nonsingular 
182/-8  (1/99)  z(FG) / Z(FG) 
183/18 (6/00)  6.30). / 6.30).  Also see Kharchenko [00] for a survey.
184/-21  (1/99)  Gabriel / P. Gabriel 
186 / 19   complement of /complement of
186/-10  (1/99) module / modular 
187 / 6  (2/99)   It can be...that / As stated in Theorem 12.4 
189 /-18,-19  (12/99) Zimmerman / Zimmermann 
189 /-9  (12/99)  Chatters and Hajarnavis [77]. / Chatters and Hajarnavis [77], who explicitly characterize indecomposable right CS rings R which are either semiprimary or have semi-primary right quotient rings as certain triangular matrices over subrings of division rings. The case when R is right Artinian is then classified (Theorems 3.1 and 6.10.) 

Furthermore: 12.5A. Theorems (Chatters and Hajarnavis [77], 
(1) A twosided Noetherian CS ring R is a pp ring with twosided maximal quotient ring. 
(2) Any twosided Noetherian CS ring has a CS Artinian quotient ring (Corollary to Theorem 6.5; Cf. Example 6.6, loc.cit.) 
(3) A right Noetherian right nonsingular right CS ring has Artinian quotient ring (Prop. 6.7. loc.cit.) 
(4) An indecomposable twosided Noetherian right nonsingular right CS ring is either prime or Artinian. 
(5) A twosided Noetherian prime ring R is twosided CS iff R is a pp ring. (Theorem 6.8; Cf. Example 6.9, loc.cit.) 
(6) A left Noetherian right PIR is twosided CS. (Corollary to Theorem 6.8, loc.cit.) Remark. The right CS property is not a Morita invariant property: there is a full 2x2 matrix ring over a right Noetherian right hereditary domain D that is not left Ore is not right CS.(Example 6.9, loc.cit.) 

190 / 20  (11/99) Replace 4. by the following: 4. If S is a submodule of a f.g projective right R-module P that is minimal with respect to S + T = P for some submodule T, then S (called a "complement submodule,"ibid.) is a direct summand of P. 
       / 25  (11/99) the next theorem. / proofs of Theorem 12.9 and the next theorem. 
190/-4,-2,-1  (1/99)  AFT / ATF 
191 /-4  (12/99)  overring / any overring 
194 / -13    Corollaries 10.14 and 10.15. / Corollaries 10.14 and 10.15. Also see Theorem 7.45. 
195/ -19 (1/01) 2.24.6B / 25.4.6B
195 /-9  (4/00) finitely embedded / finitely embedded linearly compact 
195 /-2  (2/99) ring / Noetherian ring 
196 /-12  (2/99) Cf. 13.19 / Cf. 13.19 and 13.36 
197 / 2  (2/99)  7.21ff. / 7.27ff. 
197 / 7  (12/99)  Leptin [14] / Leptin [55,57] 
197  (1/99)  To "REMARK" add: 6. Any AB-5* right R-module M over a right l.c. ring R is l.c. 
197  (1/99) To "Note" add: 4. (6), and other AB-5* results, are by the same authors [98]. 
198/10  (1/99)  3.18 / 13.18 
198 /13  (1/99)  [63] / [63b] 
199 / 3  (2/99)  13.22.4 / 13.22.4 and 13.24A 
200/ -19 (1/01) Theorem 3.5C / Theorem 3.5D
200 /-17,-18  (2/00) Add: 13.30 THEOREM (Jain [73]) Any right FGF ring R is right IF. PROOF. This is a corollary of Theorem 6.8. REMARK. Jain proved this in a different way. 
201 / 3  (2/00)  FPF / FGF 201 
        / 6,7  (3/00)  Add: REMARK. Any right coherent right FGF ring R is QF, since every cyclic right R-module R/I embeds in a free module of finite rank, hence is finitely presented, i.e., I is f.g. Thus R is Noetherian so (FGF2) applies. 
201  (2/99) In 13.32-3: Menal [82] / Menal [82b]
201 /-10  (1/00)  Pardo / Pardo and Asensio 
201 /-9  (12/99)  [97A] / [97] 
202 /1  (2/99)  [83] / [82b] 
203 / 11  (4/99)  Theorem 3.31 / Theorem 13.31 
203 /-1  (2/99)  Add: Regarding (3), cf. Theorem 7.7. 
206 /-5  (4/00)  8.12 / Theorem 3.4A 
207 / 5  (11/99)  [69a] / [69b] 
20 /-15  (2/99)  dim (m/) / dim(m/
210 /3,4 (4/00)

Add:
14/16A COHEM'S STRUCTURE THEOREM [46]. Let (R,m) be a complete regular local ring of dimension n. If R has equicharacteristic (i.e., char R = R/m), then R is isomorphic to the power series ting over R/m in n variables. (See Corollary 14.19 below. Also see Theorem 5.4.)
Proof. See, e.g. Zariski-Samuel [60],p. 307

210 / 8 (12/99)  exists / exist 
210 /-18,-19  (2/00)  Add: COHERENT POLYNOMIAL RINGS 
By a result of Soublin [68b] a polynomial ring over a coherent commutative ring R need not be coherent, but it is if R is VNR (Soublin [68a]). Cf. Carson [72]. Also see Theorem 14.21 below. 
210 /-3,-4  (2/00)  Add: REMARK. A locally coherent commutative ring R need not be coherent, but it is if R is semilocal (Harris [67]). Furthermore: (4/00) 14.21A THEOREM (Carson [78]) If R is a VNR ring of bounded index, the R[X] is coherent for any set X of commuting variables. 
212 / 18  (2/99)  See (loc.cit.) / See Gordon-Robson (loc.cit.)
213 / 14 (2/99)  dim R<. / dim R<. Cf 14.25, also 14.31A below. 
215  (2/99)  Following 14.36, add: Remark. The same result holds replacing "Radical" by "prime". See Corollary 2.4 op.cit. 
215 / 19  (12/99)  Pusat-Yilmaz, and Smith [96] / Kaplansky [74] 
215  (2/99) Following 14.38 add: Remark. The necessity of the conditions is, of course, 14.34; and the sufficiency is Exercise 25, p. 65 of Kaplansky [74]. Moreover Theorem 14.38 holds more generally for PI - rings (Pusat-Yilmaz and Smith [96]) 
217 / -10  (2/99)  Countable ordinal / ordinal 
        / -9  (2/99) ring / unique factorization domain. (Loc.cit. Corollary 9.11) 
219  (9/99)  Add: 
FACCHINI'S THEOREMS ON INJECTIVE DIMENSION 
    Facchini [82] studied rings R with the property (P) that finitely embedded right R-modules M have inj.dim. not exceeding 1; equivalently, every factor module of E(M) is injective. (It suffices that this is required of just simple modules M. (Lemma 1.4).) 
    Facchini's main theorem characterizes commutative rings R with the property (P) as locally almost maximal valuation rings whose prime ideals are either minimal or contained in a unique maximal ideal (Theorem 2.31). In case R is semilocal, then R is a direct pro- duct of almost maximal Bezout domains, and conversely, (Corollary 2.4). Moreover, R has the property (P) iff R is reduced and E(R/P) is uniserial for any non-minimal prime ideal P (Theorem 3.1). 
222 / 7  (12/99)  Corollary 5.3 / Proposition 9.5.3 
223 /-9,-10  (2/99) no coefficient...annihilates A. / when A does not have a unit element, assume that not all coefficients of f annihilate A. (See strongly regular, 15.9 and Definition 1' on p. 225.) 
225 / 18  (12/99)  Cohn [8l] / Cohn [91] 226 / 1,4 (4/99) Change the subscript on S^m to 2n. 
226  (4/99)  /
226 / 8  (2/99)  algebra k / algebra C over k 229 To "Remark 16.2", add: Cf.3.7B and C. 
230 / 1  (12/99)  Capson / Copson

/-1

(12/99) x / (x) 
231 / 15  (12/99)  right ideal of R. / right ideal of R (cf.16.9C below). 
232 /-13  (12/99) theorem / lemma 
233 / 12    Add: 
    Ass*() = Ass(M) / Ass*() = Ass*(M), in fact, the set of associated primes of a finite product of modules is the union of their associate primes. (See, e.g., Eisenbud [96], p. 93, Lemma 3.6, for a more general result.) 
233 / 21  (9/99)  Add: Cf. the Brewer-Heinzer Theorem 6.39. 
235 / 9   Indedecomposable / indecomposable 
235    16.17 Theorem is worded awkwardly. 
     Replace by: 
               16.17 Theorem. If M is a module over a commutative ring R, and if E(M) is directly decomposable, then d(M)>! Ass(M)!.
       / 17    Delete the first sentence of the proof, and delete "In the second statement," 
235 / 18   we may / We may above remark. / above definition. 
238  (2/00)  Preceding 16.33: replace the definition by :
    Definition. Let 
                  O -> M -> -> ... -> -> ... 
be a minimal injective resolution of the R-module M, and define the Noetherian depth of M, denoted n.d.M as the maximal i such that is sigma-injective V n< i. If is not sigma-injective, we let n.d.M = -1; and if is sigma-injective for all i, set n.d.M =
238 /-4  (3/99)  Beck [72] / Beck [72A] 
239  (2/99)  Preceding 16.34 insert: 16.33A BECK'S COROLLARY 3.10. A commutative ring R has n.d. >0 iff 
Q(R) is Noetherian. 
  (10/99) to 16.34 add: (4) 16.33(1) is equivalent to Q(R) being Noetherian. Thus by Camillo's theorem 16.50 (coming up), every factor ring of R satisfies 16.33(1) iff R is Noetherian. 
239 / 8   zip. / zip. (3) Cf. 3.16C. 
        / 21    Also dee / Also see 
240 / -17    R/P / R \ P 
241 /-12  (2/99)  1.10 / 16.48
242 / -9    Lam[98a] / Lam [98b] 
       / -4    Delete question (3). 
246 /-20  (12/99)  implications. there / implications; there 
248 / 18  (2/99)  4.1A applies. / 17.4A applies 
254 / 8  (12/99)  Small [86] / Small [85] 
254 /-5  (12/99)  superceded / superseded 
254 /14  (5/00) independent / dependent 
255 / 3  (1/99)  1950 /1951 
       / 12  (2/99)  about this.) / about this. Also see Bashmakova and Smirnova [99] (Amer.Math.Monthly 106 (1999)) 
for an attribution of "literal symbolism" of algebra to Diophantus (middle 3rd Century B.C.) 
255 /-10  (1/00)  name.) / name. I remember that his colleagues calling him "Al.") 
256 / 12  (12/99)  depression), / depression, 
/-1  (12/99)  Blassenohl / Blessenohl 
257 / 2  (12/99)  edged). / edged. However their methods were completely different and they were able to extend my results in the certain cases of characteristic p.) 
259 / 23  (12/99)  undergrads. / undergrads. (Not that others were not supportive. Certainly Orrin Frink, the Chair at Penn State during my two years there, maintained and supported the highest academic standards.) 
259/ 27 (2/01) chairman / chairman, Raph Hull (another of Dickson's students!),
260 / 13  (2/99)  Gilman / Gillman 
        /-17  (2/99)  10/5 
       /-16  (2/99)  mathematics. / mathematics. (In the Amer. Math. Monthly, 106 (1999) p. 97, it is stated by Kenneth A. Ross in the "Distinguished Service Award for Leonard Gillman" that "Len held a piano fellowship for five years at the Juillard Graduate School before turning to mathematics.") 
262 /-5  (1/99)  82 / 87 ("Orrin, or Frinky as I called him, was born in 1901, and died in 1988" (letter from Aline Frink of December 20, 1998).) 
262 /-18  (12/99)  Gaiété Parisiennes / Gaieté Parisienne 
263 / 9  (12/99)  Romantische / Romantisches
        / 16  (12/99)  rote / roten 
264 / 17  (12/99)  Ruhren / Beruhigen
264 /17  (1/99)  Sovort / sofort 
265 / 17  (12/99)  Mög Es / Mögen's
265 /   (1/99) insert the following additional paragraphs before "Marston Morse":

Willy, The Heidelberg VW Salesman 

Nevertheless, by looking for and too often finding flaws in the national character of Germans, I often wonder if we have not squandered a reservoir of goodwill that many Germans felt for us and the Allies for relieving them of the evident repressive evils of Nazism. Along with the bad, many of the good suffered or perished under Hitler's repugnant regime. I became good friends with the VW salesman whom I shall call Willy, who in September '59 sold me a 1953 VW "Bug" for the then magnificent sum 
of $600, or about 2400 Deutsche Marks. (The Dollar was King back then!) He told of his being "captured", i.e., surrendering to the Allies along the Rhine, and being roughly interrogated by an American who happened to be Jewish. Although Willy was fully cooperative and eager to please, the interrogator, after finding a photograph of his wife and family in Willy's wallet, tore it up in front of him. When I expressed my deepest sympathy at this bit of cruelty, he replied, "Ja, Herr Professor, but you must remember we Germans tore not merely photographs but their people to pieces!" This admission of collective guilt and the frightful imagery moved me to tears. I invited him to my home in Neuenheim, and subsequently we exchanged family visits throughout the academic year 1959-60. 

 Italiensche Reise

In Spring recess, March l959, my first wife, Mickey, and my daughters, Heidi and Cindy, two German babysitters, and I (together with our luggage under the front hood), crammed into our tiny "beetle"for a three week exploration of the South from Heidelberg to Freiberg, Basel, Zurich, the Jungfrau, Lausanne, Geneva, the French Alps, Lyon, Marseilles, Nice, Cannes, Rapallo, Pisa, Siena, Gaeta, Rome, Pompeii, Herculaneum, Ravenna, Venice, the Dolomites, Brenner Pass, Innsbruch, the Arlberg, Konstanz, Schaffhausen and back via Mannheim to Heidelberg. (Not bad for a $600 car?) One of our babysitters, tall and blond, attracted a great deal of admiration in macho Italy. Once we had to return a radio given to her as a present, when the ardent lover had his ardor doused by our curfew. Another time, when we returned from viewing the Coliseum, we saw them surrounded, again by machismo, while the children teetered on the curbside unattended with speeding cars whizzing by. ("What does not destroy you makes you stronger." -- Nietzche) 

265/-3  (1/99)  society / Society 
265 /-1     
266 / 17  (12/99) apochryphal / apocryphal 
268 / 3  (12/99)  comprised of / comprising 
269 /-10  (12/99)  Gärding/ Gårding 
270 / 2  (12/99)  in Amsterdam / at Harvard 
        / 4  (12/99)  of / off 
274 / 12  (12/99)  Adolph / Adolf 274n 
        /-4  (1/99)  later / later" 
274 /-4  (1/99)  forgot] / forgot
       /-4  (1/99)  nazi." / Nazi.] 
275 /-10  (12/99)  Schoenfliess / Schoenflies 
277 /-9, -14  (12/99)  Rudy / Rudi 
278 / 5  (12/99)  en / ein 
278/11  (1/99)  (x) / powers of x 
278 /-17 (12/99)  Rockerfeller / Rockefeller 
278 /-2  (12/99) genuises / geniuses 
279 /-2  (12/99)  apochryphal / apocryphal 
279 /-21  (12/99)  "Hörgelt" is Yiddish. The German word is Hörgeld 
280  (5/00) Wooster / Worcester
281 /-6  (3/99)  3.33 / 13.33 
281 /-1  (7/99)  To Footnote 19, add: According to Paul Halmos (in Amer. Math. Monthly, 1973, & reprinted in the World Treasury of Physics, Astronomy and Mathematics, Timothy Ferris (ed.), Little Brown (Back Bay), Boston, 1991), Maniac is an acronym for Mathematical Analyzer, Numerical Integrator, Automatic Calculator! However, the first general purpose electronic computer was the Eniac, theoretically, based on Alan Turing's "machine", or mathematical description of a machine, that in principle could solve general mathematical equations. The Eniac (=electronic numerical integrator and computer) was devised by P. Eckert and J.W. Mauchly at the University of Pennsylvania in WWII and became operational in 1946, that is, after the war. Herman Goldstine (N.B.) is the sole surviving member of the Eniac team. 
283 / 12  (1 /00)  abroad / abroad. In "Duino Elegies," Rilke has this to say about leave taking:
 "Who turned us around like this so that we always,/do what we may, retain the attitude of someone who's departing? Just as he / on the last hill that shows him all his valley / for the last time, will turn and stop and linger, / we live our lives for ever taking leave. "
283 /-7 (2/99)   E. Kolchin / Ritt 
283 .-12  (12/99)  whiskey / whisky ("whiskey" is the American and the Irish sort) 
285 /-5  (12/99) Herman / Hermann 
287 / 5  (12/99)  which / which every 
       / 11  (12/99)  at (x) / at powers of x 
287 /-20  (12/99)  17.40 ff / 7.40 ff 
287 /-17  (2/99)  John's / Johns' 
288 /-11  (12/99)  Hembda / Hemda 
289 / 1  (12/99) uncrossed / "uncrossed" (or noncrossed) 
289 / 2  (12/99) ~2 / 2.4D 
289 / 3 (12/99)  Shimson / Shimshon 
291 / -4 (2/99) A. Pedoe / D. Pedoe 
291 /-2  (12/99)  New England / England 
292 /-17  (12/99)  Chandigahr / Chandigarh 
294 / 20  (2/99) Mahabili Purim / Mahabalipuram (Seven Pagodas; also called Mamallipuram, or city of Malla) 
295 / 2  (2/99) Seven Wonders / Wonders 
295 / 21  (12/99)  Shackelton / Shackleton 
 295 / 25  (12/99)  Penquin / Penguin 
295  (2/99) Add to Footnote 25: The Seven Wonders of the World were talked for centuries B.C. (See, e.g., Leonard Cottrell, "Wonders of the World", 1959.) On the other hand, the Taj Mahal was completed in 1648. (See Columbia Encyclopedia, which asserts that the jewels on the exterior were "semiprecious stones".) 
296 / 14  (4/99)  Fisher / Fischer 
297 / 12,15  (3/99)  Nietsche/ Nietzsche 
298 / 21 (12/99) forebearer / forebear 
299 / 6  (12/99)  Rheinhold / Reinhold 
299 / -12  (3/99)  A.M.S. by Jane Scanlon / A.M.S. (Vol.11) by Jane Scanlon (Cronin) 
301/16  (1/99)  225 / 250 
301/-3  (1/99)