Greenblatt, S. H. (2003) - Harvey Cushing's Paradigmatic Contribution To Neurosurgery and The Evolution of His Thoughts About Specialization.

+DUYH\&XVKLQJ V3DUDGLJPDWLF&RQWULEXWLRQWR1HXURVXUJHU\
DQGWKH(YROXWLRQRI+LV7KRXJKWVDERXW6SHFLDOL]DWLRQ
6DPXHO+*UHHQEODWW
Bulletin of the History of Medicine, Volume 77, Number 4, Winter 2003,

pp. 789-822 (Article)
3XEOLVKHGE\-RKQV+RSNLQV8QLYHUVLW\3UHVV
DOI: 10.1353/bhm.2003.0168
For additional information about this article

http://muse.jhu.edu/journals/bhm/summary/v077/77.4greenblatt.html
Access provided by University of Bath (17 Mar 2016 11:55 GMT)

Harvey Cushing’s Paradigmatic
Contribution to Neurosurgery
and the Evolution of His Thoughts
about Specialization
S A M U E L H . G R E E N B L AT T
summary: The modern era of neurosurgery began in 1879 with the amalgam-
ation of three technologies: anesthesia, antisepsis/asepsis, and cerebral localiza-
tion. However, when Harvey Cushing (1869–1939) took his first tentative steps
toward a neurosurgical career in 1901, the outlook for the field was dismal,
because mortality and morbidity rates were horrific. For brain tumors, surgical
mortality rates were 30–50%. I will argue that Cushing made intracranial surgery
clinically effective, rather than just feasible, by adding a critical fourth technol-
ogy: knowledge and control of intracranial pressure (ICP). During his Wanderjahr
in Europe (1900–1) Cushing came to understand ICP in biophysical terms. At
Johns Hopkins, these lessons were quickly translated to acute human traumatic
cases (1901–4) and then to tumor patients with raised ICP (1903–5). By 1910,
he had accumulated enough tumor cases (180) to have convincing statistics. His
mortality rate for tumors was 10–15%. Nonetheless, the successful paradigm was
not fully instantiated until a community of practitioners formed a neurosurgical
society in 1920. As this process unfolded, Cushing’s ideas about specialization
also evolved in interesting ways.
keywords: Harvey Cushing, history, neurosurgery, intracranial pressure, brain

tumors, paradigm, specialization, translational research
The original and shorter version of this paper was presented as the Mark M. Ravitch
Lecture in the History of Medicine at the University of Pittsburgh on 9 March 2000, with the
title “Harvey Cushing’s Paradigmatic Contributions to Neurosurgery and to Specializa-
tion.” As Cushing said about one of his own presentations (“Special Field” 1920 [n. 63], p.
603), my lecture was also “made to do double duty” as the presidential address at the fifth
annual meeting of the International Society for the History of the Neurosciences, Provi-
dence, R.I., 14 June 2000. Some of the material in the opening section, “Cushing’s Place in
Neurosurgery’s Gestational Period: The Dismal Outlook circa 1900,” was originally pre-
sented as a paper, “The Dismal State of Cerebral Surgery circa 1900,” at the annual meeting
of the American Association of Neurological Surgeons, San Francisco, 12 April 2000.
I am indebted to Dale C. Smith for discussions about the substance of the section on
Cushing’s “Evolving Thoughts about Specialization, 1904 to 1920,” when we were writing
789 Bull. Hist. Med., 2003, 77: 789–822

790 samuel h. greenblatt
Among historians of neurosurgery, there is general agreement that three

essential technologies were needed to begin the modern era: anesthesia,
antisepsis/asepsis, and cerebral localization.1 There is also general con-
sensus that Harvey Cushing (1869–1939) was the single most important
figure in the ultimate success of neurosurgery as a distinct specialty.2 But
the three technologies had been available and in use for two decades
before Cushing started in 1901—so exactly what did he do that was so
crucial? I will argue that he added a fourth technology, which made
intracranial surgery clinically effective, rather than merely feasible.
By understanding raised intracranial pressure (ICP), and teaching
himself to control it, Cushing was able to bring neurosurgery’s horrific
mortality rates into a reasonable range. Thus, he made neurosurgery
successful by completing its central paradigm. But then a funny thing
happened to the paradigm: other surgeons did not understand it for
many years. Those who eventually did grasp it and use it were, for the
most part, only the small cadre who became full-time neurosurgeons,
thereby creating the community of practitioners who fully instantiated
the Kuhnian paradigm by 1920.
As Cushing went through this process, his ideas about specialization
evolved from uncertainty about what it meant, to an increasingly strong
conviction that neurosurgery must be its own independent specialty.
From the historian’s standpoint, we can use the example of Cushing and
neurosurgery to argue that new specialties may sometimes arise because
of the success of a new paradigm. That is, the emergence of a new
specialty may be primarily due to the development of a new and more
Greenblatt and Smith, “Cushing’s Leadership” (see n. 2), and I also owe many thanks to
Michael Bliss for his review of an earlier version of the manuscript, which resulted in several
very valuable suggestions. Access to archival materials (see footnotes 17, 78, and 88) was
granted and much assisted by the staffs at the Manuscripts and Archives Division of the Yale
University Library and the Alan Mason Chesney Medical Archives of the Johns Hopkins
Medical Institutions. The Bulletin’s anonymous reviewers offered cogent criticisms which I
have tried to address without lengthening the paper excessively.
1. See Samuel H. Greenblatt, “The Historiography of Neurosurgery: Organizing Themes
and Methodological Issues,” in A History of Neurosurgery in Its Scientific and Professional
Contexts, ed. Samuel H. Greenblatt, T. Forcht Dagi, and Mel H. Epstein (Park Ridge, Ill.:
American Association of Neurological Surgeons, 1997), pp. 3–9, on pp. 3–4; Robert H.
Wilkins, ed., Neurosurgical Classics (New York: Johnson Reprint Corp., 1965; reprinted Park
Ridge, Ill.: American Association of Neurological Surgeons, 1992), p. 15.
2. See Samuel H. Greenblatt and Dale C. Smith, “The Emergence of Cushing’s Leader-
ship 1901–1920,” in Greenblatt, Dagi, and Epstein, History of Neurosurgery (n. 1), pp. 167–90;
Wilder Penfield, “The Passing of Harvey Cushing,” Yale J. Biol. & Med., 1940, 12: 323–26, on
p. 325; Ernest Sachs, “The Most Important Steps in the Development of Neurological
Surgery,” ibid., 1955–56, 28: 444–50, on p. 445.
Harvey Cushing’s Paradigmatic Contribution to Neurosurgery 791
effective way of approaching an old problem, conceptually and techni-

cally, with socioeconomic factors coming into play during and after the
paradigm’s creation.
Cushing’s Place in Neurosurgery’s Gestational Period:

The Dismal Outlook circa 1900
The four decades from 1879 to 1920 were neurosurgery’s “gestational
period.” It started with William Macewen, who first used the original
three technologies to do successful intracranial procedures in Glasgow
in 1879, and it came to completion in 1919–20 with the founding of the
Society of Neurological Surgeons by Cushing and Ernest Sachs.3 Macewen’s
early efforts in provincial Glasgow went unnoticed at the time,4 but the
famous glioma operation of Alexander Hughes Bennett and Rickman
Godlee was done in 1884 in London, the capital of the British Empire. It
attracted immediate attention in the medical press and even in the Times
of London.5
3. See Greenblatt, “Historiography of Neurosurgery” (n. 1), pp. 3–5.

4. William Macewen, “Tumour of the Dura Mater - Convulsions - Removal of the
Tumour by Trephining - Recovery,” Glasgow Med. J., 1879, 12: 210–13; for later citations of
this and other cases by Macewen, see Samuel H. Greenblatt, “Cerebral Localization: From
Theory to Practice. Paul Broca and Hughlings Jackson to David Ferrier and William
Macewen,” in Greenblatt, Dagi, and Epstein, History of Neurosurgery (n. 1), pp. 137–52, on p.
152. Some authors have dated the modern era of neurosurgery to the operation of Bennett
and Godlee in 1884 (see n. 5 below), because that was the first time that a brain tumor
operation was widely recognized by the medical and general public. Nonetheless, most
historians of neurosurgery in the past half-century have recognized Macewen’s contribu-
tions: see, e.g., John E. Scarff, “Fifty Years of Neurosurgery, 1905–1955,” Surg. Gyn. Obstet.,
1955, 101: 417–513, on pp. 418–19 (reprinted in Loyal Davis, ed., Fifty Years of Surgical
Progress 1905–1955 [Chicago: Martin Memorial Foundation, 1955], pp. 303–99, on pp. 304–
5); A. Earl Walker, The Genesis of Neuroscience (Park Ridge, Ill.: American Association of
Neurological Surgeons, 1998, p. 257; Wilkins, Neurosurgical Classics (n. 1), p. 391. Arthur E.
Lyons, “The Crucible Years 1880 to 1900: Macewen to Cushing,” in Greenblatt, Dagi, and
Epstein, History of Neurosurgery (n. 1), pp. 153–66, on pp. 153–57, offers a different perspec-
tive by dating the public appearance of neurological surgery to 1881, when David Ferrier
“won” the famous debate with Friedrich Goltz about cerebral localization at the Interna-
tional Medical Congress in London.
5. The full report with discussion is in A. Hughes Bennett and Rickman J. Godlee, “Case
of Cerebral Tumour,” Medico-Chirurg. Trans., 1885, 68: 243–75. The surgery took place on
23 November 1884, and the Lancet gave a one-paragraph report (“Excision of Tumour
From the Brain”) on 29 November (p. 971). There were also subsequent short progress
reports in the Lancet on 26 December 1884 (p. 1017) and 3 January 1885 (p. 13, and
editorial on pp. 23–24), as well as a case report under the names of Hughes Bennett and
Rickman J. Godlee, “Excision of a Tumour from the Brain,” Lancet, 20 December 1884, pp.
Although Bennett and Godlee’s patient died a month after his sur-
gery, their case showed that intracranial tumors could be localized solely
by neurological examination and thence found at operation. A wave of
enthusiasm for “brain cases” began soon after Victor Horsley reported
that he had successfully removed a brain tumor in December 1886.6 John
E. Scarff’s survey of the Surgeon General’s Index Catalogue showed that
“more than 500 different general surgeons reported operations per-
formed upon the brain” in the period from 1886 to 1896; Scarff went on
to observe: “The fact that in the next 10 years from 1896 to 1906 the
number of surgeons reporting cases had fallen to less than 80 reflected
discouragement and a beginning return to sanity.”7 This discouraging
outbreak of “sanity” in the late 1890s was due to the realization that
operating on the brain was associated with terrible morbidity and mortal-
ity, even in the hands of the best surgeons of the time. Thus, when
Cushing started his career on the faculty at Johns Hopkins in 1901, the
outlook for his “Special Field”8 was not very promising, and the pessi-
mism was not merely impressionistic: there were clear statistics. Since
tumor outcomes quickly became a standard for judging the status of
brain surgery in general, I will present a small sampling of the tumor data
that were readily available at the time.
Ernst von Bergmann was one of the leaders in converting Listerian
antisepsis to aseptic technique in the 1880s and 1890s. As professor of
surgery at Berlin from 1882 to 1907, he was one of the dominant figures
in the entire world of surgery. Throughout his career, he took an active
interest in neurosurgery and neurophysiology.9 He recognized the im-
portance of raised ICP and began a series of major research publications
on the subject in 1880.10 In 1900 he published a multiauthored, multi-
1090–91. The British Medical Journal began to discuss the case on 3 January 1885 (pp. 19,
48), and the Boston Medical and Surgical Journal (now the New England Journal of Medicine)
chimed in with an editorial on 8 January 1885 (112: 41–42), titled “Vivisection or Antivivi-
section; Which Is the More Humane?” Vivisection was also the reason for the attention that
the case received in letters to the editor of the Times of London on several dates, including
16 December 1884 (p. 5), 29 December 1884 (p. 8), 3 January 1885 (p. 10), 5 January 1885
(p. 7), and 6 January 1885 (p. 10).
6. Victor Horsley, “Brain Surgery,” Brit. Med. J., 1886, 2: 670–75.
7. Scarff, “Fifty Years,” in Davis, Fifty Years of Surgical Progress (n. 4), pp. 306–7.
8. See n. 14 below.
9. See Louis Bakay, Neurosurgeons of the Past (Springfield, Ill.: Thomas, 1987), pp. 73–83;
William C. Hanigan, William Ragen, and Mary Ludgera, “Neurological Surgery in the
Nineteenth Century: The Principles and Techniques of Ernst von Bergmann,” Neurosurgery,
1992, 30: 750–57.
10. Ernst von Bergmann, Die Lehre von den Kopfverletzungen (Stuttgart: Ferdinand Enke,
1880).
volume Handbuch of surgery, which included a large volume on surgery

of the head. Bergmann wrote many of the neurosurgical chapters him-
self, based on his own experience. His mortality rate for intracranial
tumors was in the range of 19–50%, depending on the tumor’s location.11
Naturally, this experience led him to issue some conservative advice
about operating on these lesions, and his sense of caution came with the
imprimatur of the one of the world’s leading surgical clinics.
The rest of the world’s experience was consistent with Bergmann’s.
Philip Coombs Knapp was a young neurologist at Boston City Hospital
and the Harvard Medical School who began to take an interest in brain
tumors in the late 1880s, when the enthusiasm for brain cases was at its
peak. In 1899 he published an analysis of more than five hundred
cerebral tumor operations from the entire world literature.12 The overall
mortality rate was 32%; this was improved from the 46% rate that he had
found in 1889, but still formidable. Knapp also tried to determine surgery’s
“chances of benefiting the patient” by taking into account the morbidity
and mortality from all attempts to find and remove tumors, the effects of
palliative operations, and similar considerations; his statistics showed
that in 1899, 58% of the patients were not benefited by their procedures,
which was the same percentage as in 1889.13
The individual reality behind these grim statistics was captured quite
vividly by Cushing a few years later, when he described what he had seen
11. Ernst von Bergmann, “Die chirurgische Behandlung von Hirngeschwulsten,” in

Chirurgie des Kopfes, ed. Ernst von Bergmann et al., vol. 1 of Handbuch der Praktischen
Chirurgie, ed. Ernst von Bergmann, P. von Bruns, and J. von Mikulicz (Stuttgart: Ferdinand
Enke, 1900), pp. 346–52. The Handbuch was translated by W. T. Bull and W. Martin as A
System of Practical Surgery, vol. 1, Surgery of the Head (New York: Lea Brothers, 1904); in this
translation, the section on “Surgical Treatment of Brain Tumors,” pp. 317–24, omits a
paragraph in the original German edition (p. 346) that gives valuable outcome data. Also,
in the section on “Compression of the Brain and the Operations for Compression” (p.
189), the translation omits a long historical/introductory passage from the original (pp.
189–90).
12. Philip C. Knapp, “The Treatment of Cerebral Tumors,” Boston Med. & Surg. J., 1899,
141: 333–37, 359–63, 384–87. Knapp stated that he was uncertain about the exact number
of cases in his analysis, because some were probably reported more than once, and he was
not always able to sort them out (p. 334). In any case, this paper is a striking example of
premodern metanalysis, which we tend to think of as a phenomenon of the later twentieth
century. Knapp’s method of combining many cases from the literature is no different from
our current practice. The major difference is in the statistical tools that are applied to the
compilation; Knapp did not even use percentages very much, although Bergmann certainly
did in analyzing his own cases (see n. 11).
13. These percentages are taken from Scarff’s analysis of Knapp’s data: Scarff, “Fifty
Years” (n. 4), p. 420.
earlier in several different cities and countries during his years as a

medical student, a surgical resident, and a traveler in Europe:
the neurologist spends days or weeks in working out the presumable location
and nature of . . . a cerebral tumor. An operator is called in; he has little
knowledge of maladies of this nature and less interest in them, but is willing to
undertake the exploration. The supposed site of the growth is marked out for
him on the scalp by the neurologist; and he proceeds to trephine. The dura is
opened hesitatingly; the cortex is exposed, and too often no tumor is found.
The operator’s interest ceases with the exploration, and for the patient the
common sequel is a hernia, a fungus cerebri, meningitis and death.14
The main point about this sequence of deadly events is the fact that the
patient died of infection that was secondary to wound dehiscence, which
in turn was due to the raised ICP that pushed brain substance through
the bone defect left by the trephining. That is, the herniation of brain
through the trephine opening was not due simply to the local pressure of
the underlying mass; the ICP was raised throughout the intracranial
space, albeit sometimes unevenly. By 1904, when he wrote this passage,
Cushing understood this pathophysiology quite well, but most of his
contemporaries did not.
Our ability to understand and control ICP is so fundamental to our
everyday work in neurosurgery that we tend to just assume its existence—
but it has not always been there. It is my contention that Harvey Cushing
created this theoretical and practical technology in the short period
from 1901 to 1905. At the time, of course, he struggled with the problem,
but the duration of his struggle was sufficiently short that the solution
seems to have evolved seamlessly while he also worked to overcome other
difficulties, especially in the technical realm. Nonetheless, we can dissect
out some important details of Cushing’s translational endeavour by
looking carefully at several of his papers that were published between
1901 and 1905. The first phase of the process took place in Europe.
14. Harvey Cushing, “The Special Field of Neurological Surgery,” Bull. Johns Hopkins
Hosp., 1905, 16: 77–87, quotation on p. 78 (also printed with the same title in the Cleveland
Med. J., 1905, 4: 1–25). It is important to point out some nontechnical, professional aspects
of this quotation. Notice that Cushing spoke about an “operator,” not a surgeon. This was a
deliberate insult. Cushing would not dignify the fictional “operator” in this tragedy with the
title of surgeon for two reasons, which are given later on the same page: First, he felt very
strongly that the surgeon who operates on the brain should be his own neurologist, because
even the most informed neurologist could not possibly understand the surgeon’s intraop-
erative problems and the surgeon must therefore solve those problems for himself. Second,
the surgeon is likely to find those solutions only if he devotes a large portion of his time and
energy to intracranial surgery, which is exactly what Cushing was doing at the time.
Cushing’s European Research on

Intracranial Pressure, 1900–1901
Cushing completed his four years of surgical training with William S.
Halsted at Johns Hopkins in June 1900. In the latter part of his residency,
he took an increasing interest in neurosurgical cases, especially trigemi-
nal neuralgia.15 He then spent an academic Wanderjahr in Europe. Since
he deliberately visited people who were doing neurosurgery and/or
neurophysiological research, it would seem that he was looking toward
the possibility of devoting himself to this field, but “twenty years later he
remarked that he had no idea then that he would eventually specialize in
neurological surgery.”16 In any case, there is no evidence that Cushing
left America for Europe with any specific intention to work on the
problem of ICP.17
During his European tour, Cushing did his most important research
in Bern, Switzerland, with the preeminent professor of surgery Theodor
Kocher and the physiologist Hugo Kronecker.18 Like Bergmann, Kocher
also had a keen interest in brain cases and in the problem of ICP.19 He set
Cushing to work on the phenomenon that is now known to neurosurgeons
as the “Cushing reflex.” Bergmann, Kocher, and others knew that in
experimental animals the pulse slows and the blood pressure rises when
the ICP is raised. This pathophysiological reaction was presumed to
occur in humans as well, but only the pulse rate could actually be
measured.20 Accurate measurement of human blood pressure was
15. John F. Fulton, Harvey Cushing: A Biography (Springfield, Ill.: Thomas, 1946), p. 262.
16. Ibid., p. 162.
17. Fulton (ibid.) does not mention anything about this possibility. My own review of
Cushing’s letters to his father in the Cushing archives at Yale University has not revealed
anything on this subject around the time when Cushing would have been planning his trip,
or during its early stages.
18. Fulton, Harvey Cushing (n. 15), pp. 176–84, 190–93, describes Cushing’s time in
Bern in some detail. For the most extensive available discussion of the relationship between
Cushing and Kocher (and also between Halsted and Kocher), see Ira M. Rutkow, “Theodor
Kocher and His Relationship with the American Surgeons Harvey Cushing and William
Halsted,” in Theodor Kocher 1841–1917: Beiträge zur Würdigung von Leben und Werk, ed. Urs
Boschung (Bern: Hans Huber, 1991), pp. 41–51.
19. See Theodor Kocher, “Chirurgische Beiträge zur Physiologie des Gehirns und
Rückenmarks,” Deutsche Zeitschrift für Chirurgie, 1893, 35: 433–94, and 36: 1–93; idem,
Hirnerschütterung, Hirndruck und chirurgische Eingriffe bei Hirnkranheiten, in Specielle Pathologie
und Therapie, ed. H. Nothnagel, vol. 9, part 3 (Vienna: A. Hölder, 1901).
20. See Ernst von Bergmann, “Der Hirndruck und die Operationen wegen Hirndruck,”
in Bergmann et al., Chirurgie des Kopfes (n. 11), pp. 189–208 (translated in Bergmann et al.,
Surgery of the Head [n. 11], pp. 189–205); and Kocher, Hirnerschütterung (n. 19), pp. 189–201.
For Cushing’s listing of the many European investigators who were interested in ICP, see
unavailable in those days, before Cushing himself had popularized the

Riva-Rocci pneumatic cuff for this purpose. Cushing first saw the cuff in
Pavia during an Italian sojourn in March and April 1901 when he also
spent four weeks doing further work on ICP with the physiologist Angelo
Mosso in Turin.21
Cushing’s experimental procedure in Bern involved placing a sealed
glass window in a trephine hole in an animal’s skull, so he could watch
the caliber and color of the cortical vessels while the ICP was raised and
lowered. For most of the experiments, the ICP was raised by filling a soft
intracranial bag with mercury, thus creating the analog of a localized
extra-axial mass, like a subdural or epidural hematoma. By simulta-
neously recording the ICP (in the mercury bag), the blood pressure, and
the pulse rate, Cushing could demonstrate the “reflex” phenomenon of
rising blood pressure and falling pulse with rising ICP, and he could
correlate those data with the qualitative status of the cortical arteries and
veins. What he saw was a blanching of the arteries when the ICP was
raised incrementally, but the cortical anemia would improve when the
blood pressure rose. Up to very high ICPs and blood pressures the
process was reversible, but at some point the animals would die from
herniation of the brainstem through the foramen magnum if the ICP was
pushed too far.22
Harvey Cushing, “Some Experimental and Clinical Observations Concerning States of

Increased Intracranial Tension. The Mütter Lecture for 1901,” Amer. J. Med. Sci., 1902, 124:
375–400, on pp. 375–76.
21. Cushing, “Mütter Lecture” (n. 20), pp. 375–400. See also Fulton, Harvey Cushing (n.
15), pp. 190, 212–16. Michael Bliss (personal communication) has pointed out that
Cushing’s mind was well prepared to see the potential in the Riva-Rocci apparatus. When
Cushing was a resident at Johns Hopkins, he took an interest in the problem of shock,
which was a major surgical conundrum of the time. In March 1900, before he left for
Europe, he published a review of George W. Crile, An Experimental Research into Surgical
Shock (Philadelphia: Lippincott, 1899), in the Johns Hopkins Hosp. Bull., 1900, 11: 73–74, at
the end of which he said: “It is to be hoped . . . that mercurial manometers, constructed so
as to be applied to the extremities, may be employed in surgical operating-rooms for the
purpose of recording vasomotor effects and changes in blood pressure, and to indicate
impending shock more definitely than at present is possible through the medium of the
anaesthetizer’s finger on a peripheral artery.” For a thorough review of the clinical and
scientific status of studies on shock ca. 1900, see Peter C. English, Shock, Physiological Surgery,
and George Washington Crile: Medical Innovation in the Progressive Era (Westport, Conn.:
Greenwood Press, 1980), pp. 3–120.
22. Harvey Cushing, “Concerning a Definite Regulatory Mechanism of the Vasomotor
Centre Which Controls Blood Pressure During Cerebral Compression,” Bull. Johns Hopkins
Hosp., 1901, 12: 290–92; idem, “Physiologische und anatomische Beobachtungen über den
Einfluss von Hirnkompression auf den intracraniellen Kreislauf und über einige hiermit
verwandte Erscheinungen,” Mitteilungen aus den Grenzgebieten der Medizin und Chirurgie,
The experiments in Turin were done with a similar experimental model,

but there was one crucial difference. For some reason, dogs were “scarce in
Berne” but plentiful in Turin.23 Since dogs are larger than monkeys or
rabbits, it is easier to cannulate their cisterna magna. Instead of raising the
ICP by creating an artificial intracranial mass (the mercury-filled bag), in
Turin Cushing did it by pushing physiological saline into the subarachnoid
space, thereby causing a diffuse rise in ICP. This diffuse intracranial
hypertension was exerted against the entire neuraxis, including the
brainstem, more or less equally. The results were essentially similar to the
Bern experiments, but in the absence of a localized mass the animals did
not die from brainstem herniation: they died when the systemic arterial
blood pressure could not rise high enough to overcome the very high ICP.24
By the end of his time in Turin, Cushing knew that he was onto
something important and that it had clinical implications. He wrote to
his fiancée, Katherine Crowell, on 18 April 1901: “I’ve made a lucky find
in some experimental work which won’t make me famous but which will
help me and some other people understand a little better some things
about brain surgery.”25 At the same time, he quickly produced a short
paper on the subject and sent it to the Bulletin of the Johns Hopkins
Hospital, where it appeared in the issue for September 1901. It was a
summary of his experiments and conclusions from Bern and Turin. His
priority claim is stated in the second paragraph:
The fact that cerebral compression occasions a rise in blood pressure is
universally known but it does not seem to have been recognized that the
1902, 9: 773–808; idem, “Mütter Lecture” (n. 20), pp. 375–400. With regard to brain shifts
and herniation, it should be noted that Cushing did not know about transtentorial hernia-
tion of the medial temporal lobe medially across the incisural notch and into the midbrain,
which was described in the 1920s (see Samuel H. Greenblatt, “The Crucial Decade: Modern
Neurosurgery’s Definitive Development in Harvey Cushing’s Early Research and Practice,
1900–1910,” J. Neurosurg., 1997, 87: 964–71, on p. 966). Transtentorial herniation is actually
the more common clinical phenomenon, but Cushing’s experimental animals may have
generally suffered transforaminal herniation through the foramen magnum, as he said,
because the artificial masses were mostly placed centrally, near the vertex of the skull.
23. Fulton, Harvey Cushing (n. 15), p. 187. See also Cushing, “Definite Regulatory
Mechanism” (n. 22), p. 291.
24. For references to Cushing’s work in Turin, see the same papers as in n. 22. In the
“Mütter Lecture” (n. 20), p. 386, Cushing made a prescient qualitative statement of the
current formulation of cerebral perfusion pressure (CPP), which is equal to the mean
arterial blood pressure minus the ICP. For safety in our current clinical practice, we prefer
to keep that figure at a positive 50–60 mmHg or more. Cushing was saying, in effect, that a
smaller positive number is not associated with dire consequences in his experimental
animals, but at the time he had no safe way to measure ICP in humans; see text and n. 41.
25. Fulton, Harvey Cushing (n. 15), p. 188.
degree of this elevation occurs pari passu with the degree of compression . . . to
which the medullary centres are subjected. It is ordinarily stated by the
numerous experimentors who have dealt with problems of compression that
fatal symptoms originate when the intracranial pressure approaches or reaches
the height of the arterial tension. The fact that the arterial tension is a varying
quantity which regulates itself so as to overcome the effects of the increased
intracranial pressure seems never to have received attention.26
He then gave a brief review of his numerous experimental variations on

this theme, accompanied by elegantly persuasive charts of simultaneously
recorded data on blood pressure, pulse, ICP, respirations, and a time
scale. In the last paragraph, he offered a straightforward, biophysical
generalization:
As a result of these experiments a simple and definite law may be established,
namely, that an increase of intracranial tension occasions a rise of blood pressure which
tends to find a level slightly above that of the pressure exerted against the medulla. It is
thus seen that there exists a regulatory mechanism on the part of the vaso-
motor centre which, with great accuracy, enables the blood pressure to re-
main at a point just sufficient to prevent the persistence of an anaemic
condition of the bulb [medulla], demonstrating that the rise is a conservative
act and not one such as is consequent upon a mere reflex sensory irritation.27
Cushing’s pejorative remark about “a mere reflex sensory irritation”

was an expression of his dissatisfaction with the outmoded theories that
investigators at the time were using to understand ICP. In this regard,
things were in a state of muddled flux, because Bergmann and many
others had produced a wealth of experimental data, but the conceptual
framework for explaining the data was inadequate.
In essence, there were still two separate theories in the literature,
which we can label (1) vasomotor and (2) venous congestion. In the
above quotation, Cushing was referring to the idea that local vasomotor
26. Cushing, “Definite Regulatory Mechanism” (n. 22), p. 290. Fulton, Harvey Cushing
(n. 15), p. 187 n. 6, discusses Cushing’s apparent reason for sending this paper urgently to
his home Bulletin and his temerity about another publication that never existed. He also
discusses (pp. 191–93) the clash between Cushing’s American self-assurance (Selbstständigkeit)
and the niceties of publishing in the continental professorial system.
27. Cushing, “Definite Regulatory Mechanism” (n. 22), p. 292 (emphasis in original).
For a summary and assessment of this paper, see Bakay, Neurosurgeons (n. 9), pp. 80–83. In
comparison to the older contributions of Bergmann and others, Bakay credits Cushing
with (1) redemonstrating the phenomenon of the Cushing reflex by elegantly clear
experiments, and (2) giving the correct interpretation of the reflex. I now suspect that
neurosurgeons call it the “Cushing reflex” because he was the first to demonstrate the
clinical utility of the phenomenon: it can be used as an indirect way to monitor ICP.
nerves in the vascular bed of a lesion could send afferent reflex signals to
the vasomotor center in the medulla. The medullary center could then
drive the blood pressure through its control of cardiac activity and
extracranial vascular beds, such as the splanchnic.28 If this interplay went
on for too long, the entire system was thought to collapse from exhaus-
tion.29 The other pathophysiological idea that was applied to ICP was the
nineteenth-century theory of “passive venous congestion.”30 Bergmann
knew that there is an important relationship between ICP and the out-
flow of both cerebrospinal fluid (CSF) and venous blood from the head.
He postulated that passive venous congestion would obstruct CSF out-
flow, and thereby contribute to rising ICP. In this sequence of events,
death also ensued when the ICP rose to the level of the arterial blood
pressure. Even when Cushing was trying very hard to understand Berg-
mann on venous congestion, he expressed uncertainty about his own
28. The fact that the vasomotor idea was still alive in Cushing’s time is easily established
by the fact of his reference to it in the above quotation. However, finding explications of it
to cite from the literature of the time has been one of the most vexing problems in my work
on this paper. The theory goes back to the discovery of the vasomotor nerves by Charcot,
and especially by Brown-Séquard, in the 1850s: see Yves Laporte, “Brown-Séquard and the
Discovery of the Vasoconstrictor Nerves,” J. Hist. Neurosci., 1996, 5: 21–25. In the late
nineteenth century, physiologists were quite interested in the control of cerebral (and
systemic) blood flow: see Edwin Clarke and C. D. O’Malley, The Human Brain and Spinal
Cord: A Historical Study Illustrated by Writings from Antiquity to the Twentieth Century, 2d ed. (San
Francisco: Norman, 1996), pp. 793–807. Largely because of the experimental difficulties of
investigating the intracranial vasomotor system, some investigators began to deny the
existence of the vasomotor nerves: see Leonard Hill, The Physiology and Pathology of the
Cerebral Circulation: An Experimental Research (London: Churchill, 1896), pp. 76–77. Bergmann
in 1900 (“Der Hirndruck und die Operationen wegen Hirndruck” [n. 20], pp. 189–208)
touched on the function of the vasomotor center, but he said nothing directly about the
role of the intracranial vasomotor nerves. For an overview of the field from the relatively
neutral shores of America (where no research was being done), see W. T. Porter, “The
Innervation of the Blood-Vessels,” in An American Text-Book of Physiology, 2d ed., ed. William
H. Howell, vol. 1 (Philadelphia: Saunders, 1903), pp. 192–210, esp. pp. 198–99. The exact
location and physiological properties of the vasomotor receptor cells in the lower medulla
are still subjects of uncertainty and ongoing investigation; the vasomotor center is not
simply coextensive with the vagal nucleus (Anthony Marmarou, personal communication).
29. The idea of physiological exhaustion probably came from nineteenth-century
concepts of the conservation of energy. In the case of vasomotor collapse, the older
concept was very close to our current ideas about loss of cerebrovascular autoregulation,
especially in aneurysmal and traumatic subarachnoid hemorrhage. Cushing was on the
cusp of the newer idea when in 1902 he wrote about “the collapse of the vasomotor
regulatory mechanism” (“Mütter Lecture” [n. 20], p. 394).
30. See Greenblatt, “Crucial Decade” (n. 22), p. 965; G. C. Roman, “Cerebral Conges-
tion: A Vanished Disease,” Arch. Neurol., 1987, 44: 444–48.
ability to follow Bergmann’s argument.31 In hindsight, of course,

Bergmann and the others who tried to understand CSF flow in this way
could not have done it satisfactorily, because our current knowledge of
CSF circulation began to take shape only in 1913, with the work of Walter
Dandy and Kenneth Blackfan.32
As Cushing pointed out in his priority claim, his experiments showed
that the blood pressure is not merely a fixed endpoint: it varies in
correlation with the ICP. In his experimental animals, Cushing could
manipulate the ICP and/or the blood pressure to either retrieve the
animal from the brink or allow events to take their course. In other
words, in the experimental setting, Cushing could actually manage the
ICP in the clinical sense of that term. Trying to interpret these findings in
accordance with the older theories was a frustrating and unproductive
exercise.
By ignoring the older theories and adopting a more strictly biophysi-
cal approach,33 Cushing was laying the groundwork for our contempo-
rary view of ICP, which was not fully worked out until 1960.34 In Cushing’s
view, a local mass at a distance from the medullary vasomotor center can
stimulate the center simply because the continued growth eventually
causes a generalized rise in ICP, and hence a rising blood pressure,
without regard to any reflex reaction in the vascular bed that harbors the
lesion. The implication of this simplified conceptualization is that reliev-
ing generalized high ICP should reverse the rising blood pressure re-
gardless of the existence or location of a focal mass.35 As with all retro-
spective analyses, reality was less clear in prospect, so it is unlikely that
Cushing saw this clinical implication so clearly at the time. Nonetheless,
his brief remark in his letter to Katherine Crowell shows that he was able
to see at least part way down the road to clinical success before he left
Europe. Moreover, his confidence about the importance of his experi-
mental results was no doubt strengthened by the fact that Kocher eventu-
ally took them very seriously.
31. See Cushing, “Mütter Lecture” (n. 20), p. 381 n. 1. The first sentence of this
footnoted summary of Bergmann’s ideas begins: “According to the most recent view of v.
Bergmann, if we interpret it correctly” (italics added).
32. Walter E. Dandy and Kenneth D. Blackfan, “An Experimental and Clinical Study of
Internal Hydrocephalus,” JAMA, 1913, 61: 2216–17. See also Clarke and O’Malley, Human
Brain (n. 28), pp. 744–49; Wilkins, Neurosurgical Classics (n. 1), pp. 69–118.
33. See quotation above at n. 26.
34. Nils Lundberg, “Continuous Recording and Control of Ventricular Fluid Pressure
in Neurosurgical Practice,” Acta Psychiatrica et Neurologica Scandinavica, 1960, 36: Suppl. 149.
35. Obviously, removal of the mass is better if it is feasible without major damage to the
brain.
In June 1901 Cushing returned to Bern to demonstrate his experi-

mental method and results to a group of medical school professors
(Referierabend). It went well, and apparently it was at this point that
Kocher really paid attention.36 There were some tense moments when
the Germanic professor tried to force the continental system of publica-
tion on his supposed subordinate, the brash young American, but they
both kept their cool and the resulting publication was appropriately
thorough.37 Kocher also published a major monograph on head injury in
1901;38 since it made extensive use of Cushing’s results, Cushing must
have known that Kocher was going to use his data.
Translating the Experimental Pathophysiology

of ICP to Clinical Practice, 1901–1905
Four months after his return to America, Cushing gave the Mütter
Lecture at the College of Physicians of Philadelphia on 8 December
1901. His title was “Some Experimental and Clinical Observations Con-
cerning States of Increased Intracranial Tension”; the complete paper
was printed in the American Journal of the Medical Sciences in September
1902. At the beginning, Cushing said that his “personal introduction to
the subject was due to the interest of Professor Kocher,” so it is clear that
he was not thinking about ICP until he got to Bern.39
In its published form, the Mütter Lecture is the transition piece in
which Cushing gave an extensive account of his experimental results and
began their application to clinical problems.40 Initially, he talked about
the distinction between compression of the brain by a local mass and by
generalized intracranial hypertension. Much of the paper is then de-
voted to illustrating these two processes by reference to his experimental
work in Bern and Turin, with many detailed variations on the theme. For
example, if the vagus nerves and spinal cord are cut before the induction
of generalized intracranial hypertension, the blood pressure and pulse
do not change with rising ICP; but if only the vagi are cut, the blood
pressure simply follows the intracranial pressure. Cushing goes into great
detail to explain the results of each of these experimental variations in
36. Fulton, Harvey Cushing (n. 15), pp. 190–91.

37. Cushing, “Physiologische und anatomische Beobachtungen” (n. 22). See also n. 26
above.
38. Kocher, Hirnerschütterung (n. 19).
39. Cushing, “Mütter Lecture” (n. 20), p. 375.
40. The most complete report of his experimental results was given in Cushing,
“Physiologische und anatomische Beobachtungen” (n. 22).
accordance with his biophysical concept of how the reflex reaction

works, and he is persuasive (at least to this modern reader).
In the final third of his published lecture, Cushing offered clinical
examples to show how his experimental results applied to human dis-
ease. From the historical perspective, it is essential to point out that he
could not have made these correlations convincing without the accurate
blood pressure measurements that he obtained with the Riva-Rocci appa-
ratus; he made this point very explicitly in a long footnote.41 The tech-
nique gave only systolic readings, and he knew that the mean arterial
pressure is more relevant; nonetheless, the Riva-Rocci apparatus was a
reasonable substitute in humans for the arterial cannula that is used to
record blood pressure directly in experimental animals. Prior to Cushing’s
introduction of the apparatus, blood pressure was simply estimated in
the clinical setting by qualitative descriptions of the pulse, such as “thready”
or “bounding.” At Johns Hopkins, Cushing added blood pressure mea-
surements to the pulse and respiratory rate recordings on the “ether
charts” that he and Ernest Amory Codman had invented when they were
students at the Harvard Medical School in 1895.42 The charts were not
actually reproduced in the published Mütter Lecture, but they did ap-
pear in its immediate sequel.
The issue of the American Journal of the Medical Sciences for June 1903
contained another long paper by Harvey Cushing, subtitled “A Sequel to
the Mütter Lecture for 1901.”43 In the second paragraph of this paper,
Cushing repeated his rightful priority claim with regard to the variability
of the systemic blood pressure in response to changes in ICP. He pointed
out the obvious analogy between an experimentally produced intracra-
nial mass (e.g., the mercury bag) and the spontaneous or traumatic
occurrence of an intracranial (intracerebral or extra-axial) blood clot. In
these clinical situations, assuming the validity of his biophysical generali-
41. Cushing, “Mütter Lecture” (n. 20), p. 396. Harvey Cushing Society, A Bibliography of
the Writings of Harvey Cushing, 3d ed. (Park Ridge, Ill.: American Association of Neurological
Surgeons, 1993), p. 24, states that the Mütter Lecture contained Cushing’s “first reference
to blood-pressure determinations,” which he pioneered in America. His complete descrip-
tion of the instrument and its use was given in a presentation at the Boston Medical Library
on 19 January 1903 and published as Harvey Cushing, “On Routine Determinations of
Arterial Tension in Operating Room and Clinic,” Boston Med. & Surg. J., 1903, 148: 250–56.
Further details about the subsequent career of Cushing’s imported innovation are given in
Bibliography (n. 41), pp. 24–25, and especially in Fulton, Harvey Cushing (n. 15), pp. 212–16.
42. Fulton, Harvey Cushing (n. 15), pp. 93–97.
43. Harvey Cushing, “The Blood Pressure Reaction of Acute Cerebral Compression,
Illustrated by Cases of Intracranial Hemorrhage: A Sequel to the Mütter Lecture for 1901,”
Amer. J. Med. Sci., 1903, 125: 1017–44.
zation about the experimental “Cushing reflex,” the arterial hyperten-

sion “which characterizes conditions of intracranial hemorrhage, for
example, is easily accounted for, and its measured level of tension . . .
from which the medulla is suffering in consequence of the compressing
force [on the medulla].”44 In other words, for Cushing the level of the
blood pressure was an indirect indicator of ICP in the region of the
medulla. He had no way to measure ICP directly in patients, which is now
standard practice;45 nonetheless, he had an indirect monitoring system
that was far in advance of anything else at the time.
For the clinical assessment of rising ICP, the best state of the art in
Europe in 1903 was probably Kocher’s staging scheme, which was pub-
lished in 1901. Relying heavily on Cushing’s experimental results, Kocher
had postulated four pathophysiological stages of advancing cerebral
compression (Hirndruck) and three clinical stages.46 In his Mütter Lec-
ture sequel, Cushing summarized Kocher’s four pathophysiological stages
and then proceeded to explain each stage in terms of his own experi-
mental experience and his biophysical approach.47 In our contemporary
parlance, the fit was quite robust, especially for Kocher’s last two stages.
However, in his actual clinical practice in the early years at Johns Hopkins,
it appears that Cushing relied on dynamic clinical data rather than
formal staging. His published case summaries and charts show that he
monitored patients’ blood pressures and pulse rates simultaneously.
When the “Cushing reflex” became obvious and worsening, in conjunc-
tion with other signs of clinical deterioration, he determined to oper-
ate.48 Without the accuracy afforded by the Riva-Rocci apparatus, of
course, this kind of close monitoring would have been impossible.
44. Ibid., p. 1018.

45. Our current standard method of monitoring ICP measures intraventricular pres-
sure via a catheter in the frontal horn of either lateral ventricle, usually on the right. This is
anatomically supratentorial (rostral to the incisural notch that surrounds the midbrain), so
it is largely an early warning system for transtentorial herniation, which Cushing did not
know about (see n. 22). In truth, we still do not have a safe and reliable way of recording
ICP from the posterior fossa in humans, where the medulla is located. Nonetheless, the
combination of monitoring supratentorial ICP and the ready availability of computerized
imaging generally allows us to begin corrective action long before rising ICP drives the
blood pressure to the precariously high levels that Cushing was encountering.
46. Kocher, Hirnerschütterung (n. 19), pp. 186–89, 197–201. Bergmann, in “Der Hirndruck”
(n. 20), pp. 194–96, also proposed a less detailed staging scheme, based partly on his own
work and partly on that of others.
47. Cushing, “Mütter Sequel” (n. 43), pp. 1019–22. The fact that Kocher (Hirnerschütterung
[n. 19]) postulated four pathophysiological stages (pp. 186–89) and three clinical stages (pp.
197–201) can be confusing. In this “Mütter Sequel,” Cushing is clearly referring to the
pathophysiological stages, because he gives the German names for them.
48. Cushing, “Mütter Sequel” (n. 43), p. 1022.
In cases of traumatic epidural or subdural hematoma, Cushing natu-

rally advocated removing the clot if it could be located.49 However,
intracerebral hematomas are another matter, because they can be much
harder to diagnose and localize by clinical examination; nonetheless,
Cushing was able to find and evacuate a few. When he did the cranioto-
mies for any of these lesions, he generally made wide osteoplastic flaps,
so the skull piece could be left loose or simply discarded if that was
necessary in order to decompress high ICP. He was well aware that he
would raise some eyebrows by advocating surgical attacks on these diffi-
cult cases of intracerebral hemorrhage, but he felt that he had a priority
claim because he was approaching the problem with good scientific
rationale:
I am unaware that the attempt has heretofore been made with any rational
intent to bring relief to the cases of extensive hemorrhage spreading from the
capsule to the corona radiata, and I do not see any reason why we should
exclude these cases from possibilities of surgical relief simply because the
hemorrhage lies beneath the cortex. . . . The underlying or major symptoms
of [brain] compression are the same whether the collection of blood be
extradural or within the substance of the brain itself.50
Three of the five illustrative cases that Cushing published in his

Mütter Lecture sequel of 1903 had been admitted in April 1902; thus,
within six months of his appointment to the faculty at Johns Hopkins, he
had begun to apply his pathophysiological knowledge of ICP to human
cases. The final paragraph of the paper summarized Cushing’s approach
to acute intracranial hemorrhage as of late 1902:
In conjunction with other symptoms, a progressive increase in arterial pres-
sure or a high degree of the same, which has already been reached, or a
pressure which exhibits from moment to moment great alterations in level
49. In his publications of this period, Cushing generally used the term “fracture of the
base” to refer to both epidural and subdural hematomas, presumably because he associated
both types of intracranial/extra-axial hemorrhages with basal skull fractures: see Cushing,
“Mütter Lecture” (n. 20), p. 392, and “Mütter Sequel” (n. 43), pp. 1024, 1027. We still
observe the frequent correlation between temporal bone fractures (more often squamous
than basal) and epidural hematomas, but many acute subdural hematomas are not neces-
sarily associated with skull fractures.
50. Cushing, “Mütter Sequel” (n. 43), p. 1037. With computerized imaging, we can now
locate spontaneous (“hypertensive”) intracerebral hemorrhages easily and quickly, but the
indications for their surgical removal are still unsettled and controversial. The essence of
the problem is the fact that patients who suffer major damage to the basal ganglia have high
rates of morbidity and mortality, even if the clot is removed expeditiously.
may be taken as a certain indication of the advisability of early operative

intervention. In case there are localizable symptoms the site of trepanation is
plainly indicated. In case of generalized compression from widespread hem-
orrhage when there are no localizing indications, the intracranial tension
should be relieved by the elevation of a large osteoplastic flap from one
hemisphere or the other with a corresponding opening in the dura.51
The final step in Cushing’s translation of his laboratory experience to

the clinical setting was its application to elective cases, especially brain
tumors. His experiments had all been done on acute preparations,
whose similarity to rapidly expanding intracranial masses (e.g., hemor-
rhages) is readily apparent. By contrast, most brain tumors grow slowly,
so intracranial hypertension is a relatively late phenomenon in their
clinical course.52 In these chronic situations, where the blood pressure
reaction of the Cushing reflex is not seen but the ICP is high, the classical
diagnostic “clinical triad” is headache, vomiting, and papilledema
(“choked disc”).53 The surgeon who opens the cranium in these patients
is faced with essentially the same high ICP as in the more acute cases, but
there is one very important difference: the patient with a rapidly expand-
ing intracranial hemorrhage is probably dying, so desperate measures
are justifiable; brain tumor patients with the clinical triad are in misery,
but their demise is not necessarily imminent, so death at the hands of the
surgeon is not so easy to accept.
During the years 1902–5 Cushing refined his surgical technique, ap-
plying the fastidious tissue-handling of his mentor, Halsted, to the ner-
vous system.54 He also began to accumulate enough tumor cases to have
some confidence that his new approach would improve outcomes while
51. Ibid., p. 1044.

52. Cushing pointed this out in a footnote at the beginning of his “Mütter Lecture” (n.
20), p. 1017: “In a slowly forming intracranial process, as tumor growth, hydrocephalus,
etc., the [blood pressure] reactions are inconspicuous, the cerebrum accommodating itself
to the gradual increase of tension.”
53. See, e.g., Cushing, “Special Field” 1905 (n. 14), p. 79.
54. This brief statement glosses over one of the most important aspects of Cushing’s
early development as a neurological surgeon: his application of gentle Halstedian tech-
nique to the nervous system. Respect for neural tissue was a hallmark of Cushing’s tech-
nique throughout his career and a fundamental lesson that he taught to other surgeons.
See Harvey Cushing, “William Stewart Halsted, 1852–1922,” Science, 1922, 56: 461–64;
Elliott C. Cutler, “Harvey (Williams) Cushing. April 8, 1869–October 7, 1939,” Science, 1939,
90: 475–82; Penfield, “Passing of Harvey Cushing” (n. 2), p. 324. For the theme of the
present paper, this point is critical, because damaged brain tissue swells and thereby
contributes to the mass effect that raises ICP; hence, it is especially important to minimize
tissue damage in situations where the ICP is already raised.
lowering morbidity and mortality.55 In 1905 he published his first major

paper on neurosurgical technique, “The Establishment of Cerebral Her-
nia as a Decompressive Measure for Inaccessible Brain Tumors.”56 It was
clearly aimed at the surgical community, because it appeared in the first
volume of Surgery, Gynecology, and Obstetrics, which was the official journal
of the recently founded American College of Surgeons. In striking con-
trast to the papers we have reviewed so far, there is nothing in this one
about the pathophysiology of ICP or its clinical assessment in terms of the
Cushing reflex. Nonetheless, the central theme is the idea of opening the
skull (and usually the dura) in order to relieve high ICP, especially when
tumors cannot be located and/or removed.
In the first few pages of this paper, Cushing tried to explain the
papilledema, headache, and vomiting of the “clinical triad” in relation to
raised ICP. It had long been thought that papilledema was caused by an
inflammatory process in the optic nerve, but Cushing was convinced
(correctly) that it was due to raised venous pressure, because the high
ICP blocks orbital venous drainage.57 He also postulated that the head-
ache was caused by dural stretching (again, probably correctly), but he
could not really explain the vomiting. In any case, in these chronic
situations he used the clinical triad as a substitute for the Cushing reflex
to indicate the presence of raised ICP.
Having thus established the centrality of high ICP, Cushing explained
his surgical approach to its control. In the professional sense, this paper
is autobiographical, because he described in detail the two patients who
had taught him the lessons that he was trying here to teach others. The
first and more important patient was “Karl L.,” a thirty-two-year-old
machinist who ultimately proved at autopsy to have a huge, deep, left
55. See Greenblatt, “Crucial Decade” (n. 22).

56. Harvey Cushing, “The Establishment of Cerebral Hernia as a Decompressive Mea-
sure for Inaccessible Brain Tumors; With the Description of Intermuscular Methods of
Making the Bone Defect in Temporal and Occipital Regions,” Surg. Gyn. & Obstet., 1905, 1:
297–314. In his “Special Field” lecture of November 1904 (n. 14, pp. 79–80), Cushing
talked extensively about palliative decompressions for tumors and other intracranial masses,
with some references to the underlying logic of relieving ICP.
57. Early in his career on the Johns Hopkins faculty, Cushing teamed with the ophthal-
mologist James Bordley Jr. to study the optic fundus in relation to ICP. In “Establishment of
Cerebral Hernia” (n. 56), p. 298 (n. 1 in col. 1), Cushing relates: “I once, while Dr. Bordley
watched the fundus of one eye, slowly inflated a Riva-Rocci blood pressure apparatus, the
armlet of which had been placed around the neck. When a pressure of about 36 mm. of
mercury had been reached, the veins had become perceptibly dilated, and they increased
in size and tortuosity corresponding with an increase of the pressure until the latter became
so disagreeable as to lead to symptoms of faintness and nausea.” This can only be inter-
preted to mean that Cushing was strangling his own neck!
parieto-occipital tumor.58 On 8 October 1903, Cushing did a large left

parieto-occipital flap and left the bone out. The patient’s triad of symp-
toms improved rapidly, but,
as an offset to this complete alleviation of the underlying pressure symptoms,
the physiological activity in the protruding part of the brain was lost, so that
immediately following the operation there was complete visual word-blind-
ness [alexia] and complete loss of stereognosic sense in the right hand, with
greatly increased hypaesthesia of the entire right side of the body.
And this is the lesson which the case draws. A hernia protected only by the
overlying scalp is capable of such a degree of protrusion, that the tracts
radiating from the cortex must be ruptured and their function lost, even if the
[brain] oedema should not suffice to interrupt their activity.59
The remedy that Cushing proposed was the real subject of this techni-
cal paper, the “temporal intermuscular” operation.60 I have discussed its
logic and technical details in another historical paper in the neurosurgi-
cal literature.61 The main point to make here is that it was a decompres-
sion done deliberately at a distance from the intracranial mass. That
strategy makes sense only if one conceives of ICP as a generalized
intracranial phenomenon that will behave in accordance with biophysi-
cal principles. In the older, vasomotor-reflex–based conceptualization of
ICP, decompression at a distance would make less sense, because local
mass and vasomotor effects were thought to be more important. But
Cushing had already discarded the older theories. Here, in this technical
surgical paper of 1905, he was completing the translation of his labora-
tory experience to the operating room.
58. The gross appearance of the tumor in Cushing’s fig. 8, on p. 305 of “Establishment
of Cerebral Hernia” (n. 56), looks like a tentorial-falx meningioma. Cushing (p. 302) called
it a “neuroglioma,” but that was before our current classification of tumors had been
established by Cushing and his many co-workers in the 1920s and 1930s. Dr. Frank P. Smith,
emeritus professor of neurological surgery at the University of Rochester (personal com-
munication), has reviewed the pathologist’s autopsy and microscopic report of the case; he
concluded that the tumor was probably a ganglioglioma.
59. Ibid., p. 302.
60. This procedure, now called a “subtemporal craniectomy,” is used only infrequently
in current practice. The contemporary equivalent would be a partial anterior temporal
lobectomy (removal of the tip of the temporal lobe), which is performed through a
subtemporal craniectomy. In the same paper, Cushing also advocated suboccipital decom-
pression for masses in the posterior fossa (brain stem, and especially cerebellum). Again,
this procedure is not generally done any longer, because we have the diagnostic means to
see the mass and deal with it accordingly. However, even more than subtemporal craniec-
tomy, suboccipital craniectomy (decompression) remained a valid and not uncommon
strategy well into the second half of the twentieth century.
61. Greenblatt, “Crucial Decade” (n. 22), pp. 966–67.
At this very early stage in his still tentative neurosurgical career,

Cushing was rather coy about discussing his case numbers, presumably
because they were not large enough to be convincing. He did state that
he had done fifteen decompressive operations,62 but only years later did
he acknowledge that he had achieved “no instance of a really successful
tumor extirpation” in a series of twenty-nine tumors from late 1901 to
late 1904.63 On the other hand, in another publication of 1905, he said
clearly that he had had only two instances of fungus cerebri due to
wound dehiscence in his entire practice to that time.64 What this means is
that he had been faithful to his own principles from the beginning. Most
of the tumor suspects whom he was seeing were in extremis. Even if he
thought that he could locate the mass, he knew that there would be high
ICP when he opened the head, so there would be a fatal outcome if he
did not control the ICP before trying to find and remove the tumor.
Therefore he deliberately planned ahead to use decompressive opera-
tions to gain control of the ICP, rather than first looking for the tumor
and then doing a decompressive operation as an afterthought if the
exploration failed. Because of this deliberate strategy, he frequently did
one preliminary decompressive operation to control ICP before he tried
a second operation to get to the tumor, and sometimes he even did two
preliminaries.
In retrospect, this was the right idea. At the time, of course, things
were not so clear. Cushing thought that it would work, but he had not yet
proved it to himself, and certainly not to the larger world. Fortunately,
persistence and hard work paid off quite quickly: by 1910 he had done
180 tumor cases with a case mortality rate of only 13%.65 His problem
then was to try to get the rest of the world to understand and follow his
idea of controlling ICP primarily, in order to avoid the mortality and
62. Cushing, “Establishment of Cerebral Hernia” (n. 56), p. 300.

63. Harvey Cushing, “The Special Field of Neurological Surgery After Another Inter-
val,” Arch. Neurol. & Psychiatry, 1920, 4: 603–44, quotation on p. 613. The potential
discrepancy between the fifteen decompressive operations (reported in 1905) and the
twenty-nine total tumor cases in the same early period (to 1905) is explained here (on p.
613) in 1920 by Cushing’s statement that there had been “no instance of a really successful
tumor extirpation” in the twenty-nine cases that he had done up to 1905. Presumably, the
term “really successful tumor extirpation” meant to Cushing what we now mean by the
current expression “gross total tumor removal,” which implies that the surgeon cannot see
any tumor in or on the brain when he finishes.
64. Cushing, “Special Field” 1905 (n. 14), p. 80. One of these two cases of dehiscence
was apparently “Karl L.” after his second procedure.
65. Harvey Cushing, “The Special Field of Neurological Surgery: Five Years Later,” Bull.
Johns Hopkins Hosp., 1910, 21: 325–39, on p. 331. See also Greenblatt, “Crucial Decade” (n.
22), pp. 967–68.
morbidity that inevitably follows brain herniation through dehisced scalp

wounds in patients with raised ICP. In this respect, he felt that he had
made progress, albeit incomplete:
Five years ago [in 1905] . . . we had begun to feel somewhat encouraged
regarding the beneficial effects of purposeful decompressive measures. A
paper written on this subject66 was quickly followed by the citation at other
hands of numerous past cases in which trepanation had served to lessen
preëxisting symptoms; the principle, in other words, was an old one. That
relief of pressure in many instances had unexpectedly followed unsuccessful
explorations for tumors had long been a matter of general knowledge, . . . but
I am unaware that any author had previously suggested a purposeful decom-
pression irrespective of possible tumor removal . . . or had ever emphasized
the importance of decompressing at a distance from the growth as a prelimi-
nary measure should the brain be under great tension.
However this [his priority claim] may be, the decompression idea caught
hold, and the frequent examples that have come under our observation of
misguided surgery, which results in making the patient worse, seem to have
been born of an apparent misunderstanding of its most basic principles.67
In the second paragraph of the above passage, Cushing seemed to give

up on pushing his priority claim with regard to preplanned decompres-
sions. Nonetheless, he still lamented the profession’s incomplete com-
prehension of his “most basic principles”—namely, the pathophysiology
of ICP and the frequent necessity for decompressing over “silent” areas at
a distance from the suspected mass. A working comprehension of those
principles is required in order to understand the difference between
coincidental decompression when exploration fails and decompression
as deliberate strategy. Cushing was finding it difficult to get most of his
fellow surgeons to appreciate the difference because they did not under-
stand the underlying concepts.68
66. Cushing, “Establishment of Cerebral Hernia” (n. 56).

67. Cushing, “Special Field” 1910 (n. 65), p. 329 (emphasis added).
68. A possible exception to this rather gross (but valid) generalization is the team of the
neurologist William G. Spiller and the surgeon Charles H. Frazier in Philadelphia. In June
1906 they made a long presentation to the Section on Nervous and Mental Diseases of the
American Medical Association on “Cerebral Decompression: Palliative Operations in the
Treatment of Tumors of the Brain, Based on the Observation of Fourteen Cases,” pub-
lished in JAMA, 1906, 47: 679–83, 744–51, 849–53, 923–26. In Spiller’s part of the publica-
tion (on p. 682), he gave a two-line mention and a reference to Cushing’s decompression
paper of 1905 (“Establishment of Cerebral Hernia” [n. 56]). Frazier (on p. 746) gives
priority to Alfred Sänger of Hamburg for the idea of doing decompressions over “silent”
areas of the brain. Cushing had quoted Sänger’s paper on palliative operations in “Special
Field” 1905 (n. 14), p. 79, and he mentioned Sänger in the first paragraph of “Establish-
Ultimately, I suspect, many of those who did come to understand ICP

became successful neurosurgeons, often on a full-time basis, and those
who did not moved away from the field because their results were
discouraging. It is difficult to substantiate this claim in all instances, but
we can to some extent trace Cushing’s difficulty in trying to teach the
surgical community about ICP and decompression by tracing his re-
marks about it in the “Special Field” papers from 1905 to 1920. This
experience was one of the factors that ultimately led him to conclude
that neurosurgery must be its own separate specialty.
Evolving Thoughts about Specialization, 1904 to 1920

On 18 November 1904, at age thirty-five, Harvey Cushing returned to his
hometown of Cleveland, Ohio, and delivered an address on “The Special
Field of Neurological Surgery” to the Academy of Medicine of Cleveland.
It was published in early 1905.69 Fortunately for the historian, he also
delivered sequels with essentially the same titles to the same professional
society in 1910 and in 1920.70 Through these serial surveys of his chosen
field, we can follow his thoughts about the nature and scope of the
embryonic enterprise. One of the connecting threads that runs through
these quasi-autobiographical essays is ICP and decompression, including
Cushing’s efforts to teach others about them.
In his first “Special Field” lecture of 1905, Cushing did not draw any
direct connections between intracranial hypertension and specialization.
However, he was obviously thinking much about decompressive opera-
tions at this time, and he talked about their palliative value at some
ment of Cerebral Hernia” (n. 56). Frazier had part of his training with Bergmann in Berlin
(see Samuel H. Greenblatt, “Neurosurgery’s Ideals in Historical Perspective,” in Philosophy
of Neurological Surgery, ed. Issam A. Awad [Park Ridge, Ill.: American Association of Neuro-
logical Surgeons, 1995], p. 20), so he might have had the background to understand
Cushing’s technical discussions of ICP. There is no clear evidence in Frazier’s published
remarks that he really grasped Cushing’s new conceptualization of ICP, but he obviously
understood the management of ICP by decompression, because there was only one death
and one failure-to-improve in his fourteen cases. The timing and tone of Spiller and
Frazier’s paper gives me the impression that they were attempting to compete with Cushing’s
decompression paper of 1905 by trying to show that it was nothing new. Fulton, in Harvey
Cushing (n. 15), does not say that Cushing attended the AMA meeting in 1906, and Fulton
generally made note of such travels.
69. Cushing, “Special Field” 1905 (n. 14). The publication in the Cleveland Med. J. (4: 1–
25) is dated January 1905, and the publication in the Bull. Johns Hopkins Hosp. (16: 77–87) is
dated March 1905.
70. Cushing, “Special Field” 1910 (n. 65), and “Special Field” 1920 (n. 63).
length. By implication, he felt that proper use of these procedures would

contribute to the success of the nascent field. In his introduction and in
his epilogue, he proclaimed the necessity and validity of specialization.
The introduction contained an arboreal analogy that carried through
the subsequent “Special Field” papers:
I take it that we all are, or should be, branches grown from the common stem
of Medicine—in the broad Hippocratic meaning of the term—and nourished
through the roots of General Pathology. . . .
. . . it seems clear that in order to advance surgical measures . . . to deal with
diseases of the nervous system, . . . specialization, or better, concentration of
thoughts and energies along given lines is necessary. Why are we so slow to
understand that we cannot graft an independent twig of operative surgery on
any of the branches of this huge medical tree and have it grow undeformed?
Original growth from the main stem of general medicine and surgery is
necessary, but some at least of those whose inclinations follow the branch of
neurology, must do their own surgery, if it is to aid in the development of their
therapy, and not depend on the help supplied by the lukewarm assistance of
other departments.71
The analogy of medicine to a tree is clear enough here, even if its use
seems a bit tortured in some places. But an obvious question is why
Cushing should have felt so compelled to justify specialization in the first
place, especially since he was practicing it in William Halsted’s surgical
department at Johns Hopkins, where American surgical specialization
received its major impetus.72 A partial answer is given at the end of the
paper:
In talking the matter over with my surgical friends, many of them have
expressed themselves emphatically against any form of operative specializa-
tion. But, granting the wisdom and necessity of a general surgical training
beforehand, I do not see how such particularization of work can be avoided, if
we wish more surely and progressively to advance our manipulative therapy.
Are practice of hand and concentration of thought to go for nothing?73
So who were Cushing’s obstreperous “surgical friends”? It is not likely

that he is referring to his colleagues at Johns Hopkins, but there is
another group of suspects: the Society of Clinical Surgery, which held its
71. Cushing, “Special Field” 1905 (n. 14), p. 78.

72. Despite some continuing controversy about specialization in 1904–5, progress
toward our present practice patterns was fairly well along in American medicine: see
Rosemary Stevens, American Medicine and the Public Interest, rev. ed. (Berkeley: University of
California Press, 1998), pp. 43–52.
first meeting in Baltimore and Philadelphia on 13–14 November 1903.

Cushing had been one of the Society’s originators. According to John
Fulton, it was his “favorite group of surgical colleagues,” and he was very
much involved with this organization when he delivered his first “Special
Field” address in 1904.74 The Society was largely an association of general
surgeons, so presumably some of its members had taken Cushing to task
for advocating too much specialization. Since he valued the group, he
would have felt compelled to address its concerns.
In 1910, when Cushing delivered his address on “The Special Field of
Neurological Surgery: Five Years Later,” he was still concerned:
Among many of my surgical friends there exists, I am aware, a general
tendency to disparage specialization. . . . Perhaps we all should be “general
surgeons,” fitted by our comprehensive training to undertake anything in the
way of an operation which the body demands. But unfortunately this is
impossible and such an attitude often entails consequences which are not only
disastrous to the patient, but which also boomerang back unerringly to the ill
repute of surgery, in the folds of whose mantle many overventuresome opera-
tors envelop themselves.75
In the next paragraph, Cushing went on to explain his reason for advo-
cating specialization with observations that were surely autobiographical:
The real leaders of to-day [sic] in surgery owe their place not to any special
brilliancy in operative manipulations, but to their laborious experimental
investigations of certain problems of disease, whereby has been disclosed a
rational mechanical basis for a surgical therapy which can then be safely and
successfully adapted by their many followers. Lasting contributions in surgery,
as in any other field, are certain to come only as the result of such concentra-
tion.76
This is clearly a description of what Cushing himself had done in his

period of laboratory research on ICP and cerebral blood flow (1900–
74. Fulton, Harvey Cushing (n. 15), pp. 231–32. For Cushing’s recollection of how the
society was originally conceived, see ibid., p. 169. William J. Mayo was one of Cushing’s
original cofounders, and Cushing’s account agrees with the description of the same events
in Helen Clapesattle, The Doctors Mayo (Garden City, N.Y.: Garden City Publishing, 1941),
pp. 295–96. According to Clapesattle, it was actually James Mumford of Boston and George
Crile of Cleveland who called the organizational meeting of the society in New York in
1903. Crile’s autobiography (George Crile: An Autobiography, ed. Grace Crile, vol. 1 [Philadel-
phia: Lippincott, 1947], pp. 140–42), contains his version of these events, but it adds some
confusion, because he calls the same organization the “American Society of Clinical
Surgery.”
76. Ibid.
1901) and its translational application to human cases (1902–5), fol-

lowed by the initiation of a successful specialty practice using the prin-
ciples derived from the research (1905–10).77 But who were his “many
followers”? By 1910 Cushing did have a small cadre of disciples among
the research fellows and residents who worked with him at Johns Hopkins,
and, probably to a lesser extent, among some of his surgical colleagues.
There is some evidence that he was trying to teach his principles of
decompressive surgery to both groups, but mostly to his residents.
For 1908–10, the available surgical logbooks of the Johns Hopkins
Hospital show that a small number of “exploratory” or “decompressive”
craniotomies were done by one of Cushing’s surgical colleagues ( J. M. T.
Finney) and by three surgical residents (Robert T. Miller, Jr., George J.
Heuer, and John W. Churchman).78 Since the entries are quite cryptic, it
is impossible to know the nature of the lesion for any of them. However,
from some of Cushing’s remarks in his “Special Field” paper of 1910, we
can surmise that most of these cases were subdural or epidural hemato-
mas. In the body of the paper, he reviewed the status of surgical treat-
ment for various kinds of neurological lesions. The section on “intracra-
nial haemorrhages” begins with the observation that
possibly no one of the recent contributions to the surgery of the head has
proved so simple or has been so satisfactory as the decompressive measure . . .
for [extra-axial hematomas]. . . . This procedure has become more or less routine in
our clinic, and has been put upon such a basis that it may safely be regarded as a task for
the general surgeon. The same, unfortunately, cannot be said as yet of the
traumatic intracranial haemorrhages of the newborn nor for the [intracere-
bral] haemorrhages consequent upon vascular disease in the adult.79
77. See Greenblatt, “Crucial Decade” (n. 22).

78. The operating room logbooks for 1 October 1907 to 30 April 1908, and for 21 July
1908 to 11 April 1910, are in the Alan Mason Chesney Archives of the Johns Hopkins
Medical Institutions, Baltimore, Md. Unfortunately, there are no extant logbooks for the
earlier period of Cushing’s tenure on the faculty at Johns Hopkins (1901 to 30 September
1907); there are Nursing Records of Surgical Procedures and Dressings for 1901 to 1907,
but these do not include the surgeons’ names. The Chesney Archives also has the operating
room logbook for 29 January 1912 to 9 December 1913, which is of interest because
Cushing left for Harvard in late September 1912. His last recorded surgical procedure at
Johns Hopkins was a “Craniotomy” on 29 August 1912. In the period from 29 January 1912
to 29 August 1912, many neurosurgical procedures are recorded by Walter E. Dandy (18
procedures), and smaller numbers by Samuel J. Crowe (1), Emil Goetsch (2), George J.
Heuer (1), and Howard C. Naffziger (1); at this time, these five men were all residents.
Most of the procedures listed for Dandy and the others were “Craniotomy” or “Decompres-
sion.” After Cushing left, from 9 September 1912 to 31 December 1912, Heuer did eleven
procedures, mostly decompressions, and Crowe did one “Hypophysis.”
79. Cushing, “Special Field” 1910 (n. 65), p. 332 (emphasis added).
In other words, by 1910 Cushing felt that he had successfully taught

his colleagues at Johns Hopkins to correctly diagnose and operate upon
epidural and subdural hematomas (although this was not the case for
other forms of intracranial hemorrhage). Since the correct diagnosis of
epidurals and subdurals requires an understanding of ICP in order to
make the decision to operate, Cushing obviously thought that he had
achieved partial success in teaching those principles to his colleagues.
The degree of success, however, had its limits. In his preceding and
extensive discussion of brain tumors, Cushing felt that he had had some
limited success in teaching the larger medical community to diagnose
and refer tumor suspects before they were devastated, but he said noth-
ing about teaching anyone else to operate on them. Indeed, he de-
scribed tumor operations as “ticklish performances” that usually end in
disaster when done “in the hands of the inexperienced.”80
Exactly who should be doing what was very much on Cushing’s mind
at this time. On 17 May 1910 he had accepted the chair of surgery at
Harvard and its concomitant position as surgeon-in-chief at the newly
founded Peter Bent Brigham Hospital, whose construction and organiza-
tional planning had not yet begun.81 Thus, it is not surprising that his
“Special Field” paper of 1910 contains an interesting paragraph about
the best way to organize the surgical services of a teaching and research
hospital:
The ideal surgical hospital would be one whose senior appointees after a
broad general surgical training would be encouraged by continuous services
[sic] to concentrate their work on special subjects—as many subjects as there
are men who may be qualified as pathfinders. I realize, of course, that such
positions cannot be created outright and men found to fit into them. Rather,
it will happen that positions must be built around such individuals as are
available and must grow in accordance with the individual’s capabilities. A
junior staff in the meantime would carry on the general routine work, any
subdivision of which . . . may become special to-morrow [sic] through some
unexpected discovery, and again at some later day lapse back once more into
the general mill, but always, it is to be hoped, on a higher plane.82
This is a much more dynamic view of how specialties arise and decline
than we would now find realistic, but Cushing’s remarks were made
before our present system of specialty boards, residency review commit-
tees, governmental requirements, and general legalistic environment
80. Ibid., p. 331.

81. See Fulton, Harvey Cushing (n. 15), p. 339.
were in place. After all, 1910 was the year of the Flexner report.83 We
know what these things look like now, but Cushing had little inkling of
them. What he did have was the recent experience of the entire medical
community with specialization, especially at Johns Hopkins. He was
particularly interested in the situation of gynecology, perhaps because
Howard Kelly was such an important and controversial figure at Johns
Hopkins.84 Cushing observed that many advances in gynecology were
being absorbed into general surgical training and practice. He was
especially concerned that some men who had trained as gynecologists
were finding the field “worked out” in terms of further advances, and so
“they unconsciously become again general surgeons—many of them,
alas, with their general surgical knowledge built upon the sands of
inexperience.”85 Obviously, he wanted to protect his own “Special Field”
from such depredations.
In tracing the history of a single man and his thoughts, it is a long jump
from 1910 to 1920, because the world changed so much over that decade.
The catalyst for this change, of course, was the Great War of 1914–18. Its
advent certainly affected Cushing’s life. He arrived in Boston from Balti-
more in September 1912, but the new Brigham Hospital did not open
fully until the spring of 1913. The war started in Europe in 1914, and
Cushing volunteered enthusiastically to serve with the Ambulance
Américaine military hospital in France for three months in 1915. After the
United States entered the war in 1917, Cushing went back to France with
Harvard’s Base Hospital No. 5 in May 1917. He returned to America in
February 1919. On balance, he felt that neurosurgery had lost more than
it gained from the experience of the war.86 Personally, he had probably
learned more than he wanted to know about military bureaucracy and its
83. Abraham Flexner, Medical Education in the United States and Canada: A Report to the
Carnegie Foundation for the Advancement of Teaching (New York: Carnegie Foundation, 1910).
84. For a sketch of how Kelly was perceived in the Johns Hopkins community, see
Michael Bliss, William Osler: A Life in Medicine (Toronto: University of Toronto Press, 1999),
pp. 215–16.
85. Cushing, “Special Field” 1910 (n. 65), p. 328. Three years later, in a major “Address
in Surgery” at the Seventeenth International Congress of Medicine in London, Cushing
expressed a similar sentiment in a more positive way: “it would be rejuvenating to many of
the surgical specialties if they could be periodically absorbed by general surgery, to be born
again under the leadership of individuals who through fresh contributions could give a new
impulse to a senescent subject” (“Realinements [sic] in Greater Medicine: Their Effect
upon Surgery and the Influence of Surgery upon Them,” Lancet, 9 August 1913, 2: 369–75,
quotation on p. 373).
86. Harvey Cushing, “Neurological Surgery and the War,” Boston Med. & Surg. J., 1919,
181: 549–52.
organization of specialty care.87 He had also learned some important

lessons about trying to teach general surgeons to do neurosurgery.
On 21 October 1919, Cushing presented a paper on brain tumor
statistics to a meeting of the American College of Surgeons in New York.
Unfortunately, the full text of the paper was never published, so we have
only a medical reporter’s summary and some of the manuscript in the
Cushing archives at Yale.88 Cushing’s emphasis on this occasion was on
tumor classification and intraoperative findings, but we know from other
sources that his case mortality rate for his first 130 tumor patients in
Boston was 8.4%.89 By September 1920, he had done more than a
thousand tumor cases altogether.90 Something about the 1919 presenta-
tion was surely impressive. When Cushing finished, the chairman of the
session, William J. Mayo, “rose and solemnly announced: ‘Gentlemen: we
have this day witnessed the birth of a new specialty—neurological sur-
gery.’”91 As a direct result of this event, the Society of Neurological
Surgeons was founded in 1920, on the model of the Society of Clinical
Surgery. Thus, Cushing’s “Special Field” paper of 1920 provides an
insight into his thoughts about specialization just when his own efforts
were coming to fruition.
The first section of the “Special Field” paper of 1920 deals extensively
with “Specialization in General.” Cushing repeated his idea about spe-
cialties arising and then returning to their “parents,” after “a period of
more or less disrepute” because of diminishing quality in the “imitators
of the pioneers.”92 Then, in the next paragraph, he continued his arbo-
real analogy with a delightfully creative variation:
87. Ibid., pp. 551–52; Harvey Cushing, From A Surgeon’s Journal (Boston: Little, Brown,
1936), pp. 357, 382. Because he sent his wife a harsh criticism of a British surgeon, Cushing
was nearly court-martialed in May 1918: see Fulton, Harvey Cushing (n. 15), pp. 426–28.
88. Harvey Cushing, “Brain Tumor Statistics,” Med. Rec., New York, 6 March 1920, 97:
417–18; Cushing’s Bibliography (n. 41), p. 40, states that this was “A stenographer’s note of
remarks at the Clinical Congress of the American College of Surgeons held in New York,
October 21, 1919. Paper never published.” In the Cushing Archives at Yale University, New
Haven, Conn., Series III, folder 133 (reel 143, box 179), there is a large collection of
material that is labeled “Brain tumor statistics (179) 1920,” but it is not a coherent
manuscript; rather, it is pieces of a manuscript interspersed with notes about individual
cases and handwritten compilations of statistics. The short report in the Medical Record is
more useful as a record of what Cushing actually said.
89. Harvey Cushing, “Concerning the Results of Operations for Brain Tumor,” JAMA,
1915, 64: 189–95, on p. 193. The operative mortality rate for 149 procedures on these 130
patients was 7.4%.
91. Ernest Sachs, Fifty Years of Neurosurgery: A Personal Story (New York: Vantage Press,
1958), p. 68.
Medicine has grown in the fashion of a banian [sic] tree. In the beginning
there was a single stem. This, in the remote past, each professional aspirant
learned to climb, and in his lifetime could encompass without difficulty all the
knowledge its various branches represented. But from the main branches
which the original tree put out, particularly from physic and surgery repre-
senting the application of medical knowledge to practice, there soon dangled
many roots. Some of these finally reaching the ground became attached, and
drawing up their own nourishment have in some cases enlarged and become
permanent and necessary supports to the parent branches.93
The last sentence appears to certify a degree of permanence for some

specialties, but they are unnamed. Presumably Cushing counted neuro-
surgery among them, because he considered it to be solidly grounded in
the basics, unlike some specialties that arose almost accidentally.
In his “Special Field” paper of 1910, Cushing had remarked that “a
new [specialty] department may any day be justified by an unexpected
discovery which at the time may seem trifling,” and he went on to name
Chamberlen’s obstetrical forceps, the Wagner cranial flap, Helmholtz’s
“optical toy” (the ophthalmoscope), Babington’s mirror for laryngos-
copy, and others.94 In his 1920 paper, he repeated the same sentiment
and named as examples “aviation,” the Wassermann reaction, and
“Roentgen’s ray,” among others.95 But then, as if to protect the integrity
of neurosurgical specialization, which he was about to discuss, he re-
flected that novelties like these
must be regarded as accidental occasions for specialization, to be distin-
guished perhaps from the purposeful determination to specialize within an
established branch long occupied. But under both circumstances those who
can best take advantage of existing opportunities, or can originate others, not
only must have a good general training in clinical medicine and surgery, but
must have been thoroughly schooled in the fundamental subjects—in the
anatomy, physiology and chemistry of morbid as well as of normal tissues and
organs—for without this knowledge any special branch is supported by a root
lodged in sand which does not long survive overloading. There is only one way
to get a secure seat on any outlying branch and that is by approaching it from
the main stem, no matter how wearisome, laborious and time-consuming this
process may be.96
Obviously, Cushing felt that he himself—and others of his genera-

tion—had endured that laborious process, so neurosurgery was on a
93. Ibid.
96. Ibid., pp. 605–6.
sound footing. Indeed, the war had confirmed his conviction about the
importance of thorough training in the fundamentals. Under the pres-
sure of wartime circumstances, short courses for training neurosurgeons
had been set up in New York, Philadelphia, and St. Louis. Cushing
remarked that these efforts had served
to concentrate attention on . . . [neurosurgery] and to give certain medical
officers, whose inclinations lay in this direction, some general ideas of trau-
matic neurosurgery.
Excellent as all this was, it is a far cry from traumatic to pathologic surgery,
from doing a peripheral nerve suture to a trigeminal neurectomy, from the
repair of a cranio-cerebral injury to the removal of a brain tumor.97
One of the major differences between debriding an open brain wound

and operating on a tumor, of course, is the problem of managing the
high ICP that is associated with the presence of the tumor mass. By 1920,
Cushing had clearly despaired of teaching the principles of decompres-
sive craniectomies to the entire general surgical community. He said that
he still considered subtemporal and suboccipital decompressions to be
“the two most useful procedures in craniocerebral surgery,” though they
were not perfected and “there are right and wrong ways of performing
them.”98 To this statement he then added a telling footnote:
I would have felt ere this that the principles of a subtemporal decompression
were so well understood that it was a safe measure in the hands of every
general surgeon, but we see so many sorry results of so-called decompression
operations . . . that the happy time does not seem to have come as yet.99
This is the same lamentation that he had sung in the “Special Field”
paper of 1910,100 but here with a tone that implied a more fatalistic
acceptance of this sorry fact of life.
The bulk of the “Special Field” paper of 1920 was devoted to thorough
reviews of Cushing’s major interests and results in neurosurgery at the
time—technique (including decompression), tumors, pituitary disor-
ders, cerebrospinal fluid, the spinal cord, and peripheral nerves. In
contrast to the state of the same interests in 1905, and even in 1910, the
97. Ibid., p. 634. Since most of the transcranial injuries in World War I were due to low-
velocity missiles (bullets, shrapnel), their emergent neurosurgical care would have con-
sisted of debridement without decompression. Acute rises in ICP would have occurred
later, due to cerebral edema in the tract of the missile or the appearance of cerebral abscess
or meningitis.
98. Ibid., p. 609.
99. Ibid.
progress had been enormous, although most of the work still lay ahead.
Part of what finally led Cushing to see neurosurgery as a freestanding
enterprise was the necessity for thorough grounding in general surgery
and neurology for the next generation of neurosurgeons, which he was
then training. That kind of investment by an individual could be justified
only by full-time devotion to the field. Thus, the overall tone of this paper
in 1920 is acknowledgment of the necessity for neurosurgery to be its
own specialty, because it must be its own master. Nonetheless, at the very
end of the paper, Cushing’s final words on the matter are counterintuitive
and nostalgic: “But for its own good, I pray that neurologic surgery may
never get so far from the home of general Medicine and its immediate
parent, surgery, that there will be any estrangement or any possibility of
its being shut out of doors when the time comes for its return.”101
Despite the wistful tone of this passage, Cushing understood that he
had launched an independent entity. With a sense of personal parent-
hood, he was reminding his offspring not to stray too far from the solid
foundations of its original home.
Enduring Issues about Specialization:

The Kuhnian Paradigm as Medical Historiography
In my view, neurosurgery completed its gestational period in 1920. It was
up and running as a distinct specialty. As such, it had its own successful
knowledge and techniques, it had a professional organization that em-
braced both full-time and part-time practitioners, and it had the recogni-
tion of the larger medical and surgical community.102 Needless to say, this
achievement was not solely Cushing’s—by 1920 he had many North
American and European colleagues—but he had clearly earned the
leadership of the “Special Field” through his efforts to create it and to
teach its principles to others. The historical dynamics of the process of
this specialization can be explored by putting it in the context of Thomas
Kuhn’s paradigm, which also helps to define some historiographic limita-
tions on our perceptions of the process.
My understanding of a complete and successful Kuhnian paradigm is
that it consists of (1) a model/method of problem-solving which has (2)

102. In my introductory chapter to Greenblatt, Dagi, and Epstein, History of Neurosurgery
(n. 1), p. 3, I proposed a set of criteria for defining a profession. When that was written, I
had no conscious intention to derive the criteria from any Kuhnian standard, but the
similarity between those criteria and my present understanding of Kuhn’s paradigm (as
expressed in the next paragraph of this paper) now seems quite striking.
both theoretical and practical features that (3) are shared by a commu-
nity of scientists/practitioners who are all interested in the same prob-
lem or set of problems.103 Since medicine is always deeply embedded in
its socioeconomic and cultural environment, we must add (4) accept-
ance of the legitimacy of circumscribed paradigms (i.e., specialties/
subspecialties) by the larger community of scientists/physicians, and
hence by the general society. Given the notorious ambiguity of the
paradigm concept, I see nothing in Kuhn’s description that would pre-
clude such an addition. Indeed, the existence of a subcommunity of
qualified scientists/physicians who share a paradigm about a delimited
problem would usually imply the recognition of their existence by their
larger community, and hence by the larger society. When funding is
involved, that external recognition generally has to be formal and
structured.
With regard to (1) a method/model of problem-solving, Cushing’s
investigation of ICP and his successful method of solving the clinical
problem with decompression surely fills the bill. Today we generally
control ICP with osmotic agents, direct ventricular drainage, and hyper-
ventilation. Those techniques have been available only in the last several
decades, but their use is still guided by our understanding of ICP, which
we now monitor directly at the bedside, especially in trauma cases. When
we prepare to open a craniotomy for tumor, we pretreat for cerebral
edema with steroids and osmotic agents until we are sure that we will not
face an uncontrollably herniating brain when the dura is opened. And
yes, we still do occasional decompressive operations, albeit infrequently.
Hence (2), the pathophysiology of ICP is still a central part of the
theoretical underpinning of the practice of neurosurgery, along with
cerebral (and all other neurological) localization. The principles of ICP
and its management are drilled into residents at the start of their train-
ing, because this knowledge is fundamental to successful practice and to
any greater level of sophistication.
In Kuhn’s original discussion of the paradigm, it is clear that the
paradigm has practically no existence outside of (3) the community of
103. This is my interpretation of Kuhn’s use of the term “paradigm” throughout

Thomas Kuhn, The Structure of Scientific Revolutions, 2d ed. (Chicago: University of Chicago
Press, 1970), and in idem, “Second Thoughts on Paradigms” (1974), reprinted in idem, The
Essential Tension: Selected Studies in Scientific Tradition and Change (Chicago: University of
Chicago Press, 1977), pp. 105–26. Despite the controversy and misunderstandings that the
word “paradigm” has evoked, Kuhn did not really elaborate on it in his later writings; see
Thomas Kuhn, The Road Since Structure: Philosophical Essays, 1970–1993, with an Autobio-
graphical Interview, ed. James Conant and John Haugeland (Chicago: University of Chicago
Press, 2000), where it is seldom mentioned.
people who share it; he does not elaborate very much on the characteris-
tics of the community, but it must be there.104 In 1910 there were many
North American surgeons who were interested in neurosurgery, but it is
hard to say that they were any kind of coherent community.105 By 1920,
there was a formal organization of surgeons who took a special interest in
the nervous system, the Society of Neurological Surgeons. The immedi-
ate impetus to its existence was William Mayo’s spontaneous proclama-
tion of neurosurgery’s birth at the meeting of the American College of
Surgeons in 1919.106 Here is an obvious example of a leading member of
a large specialty (surgery) acknowledging the legitimacy of a subspecialty
in its midst, thus fulfilling criterion (4) for the existence of a full para-
digm in the expanded Kuhnian sense.
The question/objection that will immediately arise is whether all
specialties must fulfill these Kuhnian criteria in order to be legitimate. To
this the obvious answer can only be, not always. Medical and surgical
subspecialties come into existence for many reasons. Some of those
factors can be sorted out and understood by looking at them through
other historiographic lenses, while keeping the Kuhnian model in mind.
The socioeconomic and professional reasons for specialization have
been investigated frequently and thoroughly in the historical and socio-
logical literature since George Rosen published The Specialization of Medi-
cine with Particular Reference to Ophthalmology in 1944.107 Although Rosen
placed the origin of ophthalmology in Helmholtz’s invention of the
ophthalmoscope in 1851, his emphasis was actually on the broader
historical and social factors in the specialty’s instantiation. In 1971,
Rosemary Stevens concluded that “specialization is the fundamental theme
for the organization of medicine in the twentieth century.”108 She and
many others have articulated the various factors that led to the rise of
104. I think it is this communal aspect of the paradigm idea that appeals so much to
physicians and others who are themselves participants in such subcommunities; e.g., see
John R. Absher and D. Frank Benson, “Disconnection Syndromes: An Overview of
Geschwind’s Contributions,” Neurology, 1993, 43: 862–67.
105. See Greenblatt and Smith, “Emergence of Cushing’s Leadership” (n. 2), pp. 179–
80.
106. See text above at n. 91.
107. George Rosen, The Specialization of Medicine with Particular Reference to Ophthalmology
(New York: Froben Press, 1944). Rosen was explicitly undertaking “a sociological study of
specialization in medicine” (p. 3), albeit with a strong historical perspective, since his
doctoral thesis was written in Columbia University’s Department of Sociology (p. 93).
108. Stevens, American Medicine (n. 72), p. ix (emphasis in original). This quotation is
actually Stevens’s summary in 1998 of her book that was originally published in 1971 and
“updated” in 1998.
specialization in the United States.109 The particular analysis that best fits
Cushing’s historical situation was given by William Rothstein in 1972:
Specialization could not develop in medicine until a number of conditions
were fulfilled: (1) a medically valid body of medical knowledge and tech-
niques had to develop in a given specialty; (2) urban population aggregates
had to become sufficiently large to support a specialist in the practice of his
specialty; and (3) institutions and arrangements within the profession had to
make it financially rewarding for a physician to restrict his practice to a
specialty.110
Like Rothstein’s, all of the sociohistorical analyses include the obvious

factor of a body of specialized knowledge that is central to the events, but
practically no attention has been given to the shape and content of that
specialized knowledge. I submit that the case of Cushing and neurosur-
gery shows that socioeconomic forces alone are not sufficient to beget a
specialty that endures: the body of specialized knowledge must have
sufficient substance and validity to hold together over time.
In this regard, I would like to suggest that neurosurgery’s example is
also relevant to some of today’s discussions about subspecialization,
which is being hotly debated in neurosurgery and in other “special
fields.” The current arguments are usually framed in economic, social, or
even political terms, but we should remember that successful specialties
may be based on successful paradigms, whose existence and intellectual
content may be germane to the contemporary discussion. If a subspecial-
izing group wants to separate itself from or within an existing specialty
because of socioeconomic factors, there may be some room for give and
take. However, if a new community of practitioners is forming in re-
sponse to its own emerging paradigm, then schism may be inevitable and
perhaps for the best. It behooves the participants in today’s debates to
recognize the difference.
109. The “various factors” usually boil down to three: (1) the existence of a body of
knowledge beyond the competency of the general practitioner; (2) the growth of hospitals
as centers of specialized practice; and (3) the explosion of medical knowledge due to
continuing research. These basic themes have recurred in the analyses of multiple scholars,
presumably because they are valid observations. See, e.g., W. F. Bynum, Science and the
Practice of Medicine in the Nineteenth Century (Cambridge: Cambridge University Press, 1994),
p. 193; William G. Rothstein, American Physicians in the Nineteenth Century: From Sects to Science
(Baltimore: Johns Hopkins University Press, 1972), p. 14; Stevens, American Medicine (n.
72), pp. ix–x.
110. Rothstein, American Physicians (n. 109), p. 207.

Greenblatt, S. H. (2003) - Harvey Cushing's Paradigmatic Contribution To Neurosurgery and The Evolution of His Thoughts About Specialization.

Uploaded by

Copyright:

Available Formats

You might also like

Greenblatt, S. H. (2003) - Harvey Cushing's Paradigmatic Contribution To Neurosurgery and The Evolution of His Thoughts About Specialization.

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Greenblatt, S. H. (2003) - Harvey Cushing's Paradigmatic Contribution To Neurosurgery and The Evolution of His Thoughts About Specialization.

Uploaded by

Copyright:

Available Formats

+DUYH\&XVKLQJ V3DUDGLJPDWLF&RQWULEXWLRQWR1HXURVXUJHU\

Bulletin of the History of Medicine, Volume 77, Number 4, Winter 2003,

For additional information about this article

Access provided by University of Bath (17 Mar 2016 11:55 GMT)

keywords: Harvey Cushing, history, neurosurgery, intracranial pressure, brain

789 Bull. Hist. Med., 2003, 77: 789–822

Among historians of neurosurgery, there is general agreement that three

effective way of approaching an old problem, conceptually and techni-

Cushing’s Place in Neurosurgery’s Gestational Period:

3. See Greenblatt, “Historiography of Neurosurgery” (n. 1), pp. 3–5.

volume Handbuch of surgery, which included a large volume on surgery

11. Ernst von Bergmann, “Die chirurgische Behandlung von Hirngeschwulsten,” in

earlier in several different cities and countries during his years as a

Cushing’s European Research on

unavailable in those days, before Cushing himself had popularized the

Harvey Cushing, “Some Experimental and Clinical Observations Concerning States of

The experiments in Turin were done with a similar experimental model,

He then gave a brief review of his numerous experimental variations on

Cushing’s pejorative remark about “a mere reﬂex sensory irritation”

ability to follow Bergmann’s argument.31 In hindsight, of course,

In June 1901 Cushing returned to Bern to demonstrate his experi-

Translating the Experimental Pathophysiology

36. Fulton, Harvey Cushing (n. 15), pp. 190–91.

accordance with his biophysical concept of how the reﬂex reaction

zation about the experimental “Cushing reﬂex,” the arterial hyperten-

44. Ibid., p. 1018.

In cases of traumatic epidural or subdural hematoma, Cushing natu-

Three of the ﬁve illustrative cases that Cushing published in his

may be taken as a certain indication of the advisability of early operative

The ﬁnal step in Cushing’s translation of his laboratory experience to

51. Ibid., p. 1044.

lowering morbidity and mortality.55 In 1905 he published his ﬁrst major

55. See Greenblatt, “Crucial Decade” (n. 22).

parieto-occipital tumor.58 On 8 October 1903, Cushing did a large left

At this very early stage in his still tentative neurosurgical career,

62. Cushing, “Establishment of Cerebral Hernia” (n. 56), p. 300.

morbidity that inevitably follows brain herniation through dehisced scalp

In the second paragraph of the above passage, Cushing seemed to give

66. Cushing, “Establishment of Cerebral Hernia” (n. 56).

Ultimately, I suspect, many of those who did come to understand ICP

Evolving Thoughts about Specialization, 1904 to 1920

length. By implication, he felt that proper use of these procedures would

So who were Cushing’s obstreperous “surgical friends”? It is not likely

71. Cushing, “Special Field” 1905 (n. 14), p. 78.

ﬁrst meeting in Baltimore and Philadelphia on 13–14 November 1903.

This is clearly a description of what Cushing himself had done in his

1901) and its translational application to human cases (1902–5), fol-

77. See Greenblatt, “Crucial Decade” (n. 22).

In other words, by 1910 Cushing felt that he had successfully taught

80. Ibid., p. 331.

organization of specialty care.87 He had also learned some important

The last sentence appears to certify a degree of permanence for some

Obviously, Cushing felt that he himself—and others of his genera-

One of the major differences between debriding an open brain wound

Enduring Issues about Specialization:

101. Cushing, “Special Field” 1920 (n. 63), p. 637.

103. This is my interpretation of Kuhn’s use of the term “paradigm” throughout

Like Rothstein’s, all of the sociohistorical analyses include the obvious