3d1405139879_218213b8c72c0a6107510b50bdde76d3

.Earth-Science Reviews 54 2001 321348www.elsevier.comrlocaterearscirev

Sequence stratigraphy as a scientific enterprise: the evolution andpersistence of conflicting paradigms

Andrew D. Miall a,), Charlene E. Miall ba Department of Geology, Uniersity of Toronto, 22 Russell St., Toronto, ON, Canada M5S 3B1

b Department of Sociology, McMaster Uniersity, Hamilton, ON, Canada L8S 4M4Accepted 21 November 2000

Abstract

In the 1970s, seismic stratigraphy represented a new paradigm in geological thought. The development of new techniquesfor analyzing seismic-reflection data constituted a Acrisis,B as conceptualized by T.S. Kuhn, and stimulated a revolution instratigraphy. We analyze here a specific subset of the new ideas, that pertaining to the concept of global-eustasy and the

wglobal cycle chart published by Vail et al. Vail, P.R., Mitchum, R.M., Jr., Todd, R.G., Widmier, J.M., Thompson, S., III, .Sangree, J.B., Bubb, J.N., Hatlelid, W.G., 1977. Seismic stratigraphy and global changes of sea-level. In: Payton, C.E. Ed. ,xSeismic StratigraphyApplications to Hydrocarbon Exploration, Am. Assoc. Pet. Geol. Mem. 26, pp. 49212. The

.global-eustasy model posed two challenges to the Anormal scienceB of stratigraphy then underway: 1 that sequencestratigraphy, as exemplified by the global cycle chart, constitutes a superior standard of geologic time to that assembled from

.conventional chronostratigraphic evidence, and 2 that stratigraphic processes are dominated by the effects of eustasy, to theexclusion of other allogenic mechanisms, including tectonism.

While many stratigraphers now doubt the universal validity of the model of global-eustasy, what we term theglobal-eustasy paradigm, a group of sequence researchers led by Vail still adheres to it, and the two conceptual approacheshave evolved into two conflicting paradigms. Those who assert that there are multiple processes generating stratigraphic

.sequences possibly including eustatic processes are adherents of what we term the complexity paradigm. Followers of thisparadigm argue that tests of the global cycle chart amount to little more than circular reasoning. A new body of workdocumenting the European sequence record was published in 1998 by de Graciansky et al. These workers largely follow theglobal-eustasy paradigm. Citation and textual analysis of this work indicates that they have not responded to any of thescientific problems identified by the opposing group. These researchers have developed their own descriptive andinterpretive language that is largely self-referential.

Through the use of philosophical and sociological assumptions about the nature of human activity, and in particular the .work of Thomas Kuhn, we have attempted to illustrate 1 how the preconceptions of geologists shape their observations in

.nature; 2 how the working environment can contribute to the consensus that develops around a theoretical approach with a .concomitant disregard for anomalous data that may arise; 3 how a theoretical argument can be accepted by the geological

.community in the absence of AproofsB such as documentation and primary data; 4 how the definition of a situation and the .use or non-use of geological language AtextsB can direct geological interpretive processes in one direction or another; and 5

) Corresponding author. Tel.: q1-416-978-3022; fax: q1-416-978-3938. .E-mail address: [email protected] A.D. Miall .

0012-8252r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. .PII: S0012-8252 00 00041-6

( )A.D. Miall, C.E. MiallrEarth-Science Reiews 54 2001 321348322

how citation patterns and clusters of interrelated Ainvisible collegesB of geologists can extend or thwart the advancement ofgeological knowledge. q 2001 Elsevier Science B.V. All rights reserved.

Keywords: paradigm; global; stratigraphy

1. Introduction

The business of research is new ideas. Fewthings, howeer, are more unpopular with re-

searchers than truly new ideas Vail, 1992, p..83 .

Emerging in the 1970s, seismic stratigraphy, as . .developed by Vail 1975 and Vail et al. 1977 at

Exxon, brought about a major revolution in the studyof stratigraphy. It re-energized a discipline, 200 yearsin the making, offering new theoretical possibilitiesfor knowledge of earth history and fundamental earthprocesses. It also provided a potentially powerfulnew analytical and correlation tool for use by prac-ticing basin analysts, especially those in thepetroleum industry. Among the new concepts andmethodologies constituting seismic stratigraphy werethe following:

.1 The use of a new form of datareflection-seismic recordsfor the generation of stratigraphicinformation;

.2 Demonstration that the new data could provideimages of large swaths of a basin at once;

.3 Demonstration of the complex internal archi-tecture of basin fills;

.4 Demonstration that stratigraphic successionsconsist of Asequences,B which are packages of con-formable strata bounded by regional unconformities;

.5 The proposal that the bounding unconformitiesare mostly global in extent and were generated byrepeated eustatic changes in sea level. We refer hereto this hypothesis as the global-eustasy model;

.6 The proposal that the earths stratigraphicrecord, consisting of a global record of sequences,could be characterized by a global cycle chart,which could be used as a universal correlation tem-plate.

Within the framework of scientific revolutions .established by Kuhn 1962, 1996 , seismic stratigra-

phy could be said to constitute a new paradigm. .Kuhn 1962 defined paradigms as A . . . universally

recognized scientific achievements that for a time

provide model problems and solutions to a commu-nity of practitioners.B These new developments haveproved to be of profound importance to the scienceof stratigraphy. Building on points 1 to 4, stratigra-phers have developed an entirely new way of practic-ing their craft, including application of the conceptsto outcrop and subsurface well data, in a new science

termed sequence stratigraphy Posamentier et al.,1988; Posamentier and Vail, 1988; Van Wagoner et

.al., 1990 . In this science, the architecture and pre-dictability of sequences are among its most valuablecomponents.

In the early days of seismic stratigraphy, in thelate 1970s and early 1980s, the global cycle chartwas considered an inseparable part of the newmethodology. The 1970s were a period when global-ization was a theme in many facets of scientific andsocietal development. Marshall McLuhan was teach-ing us about a Aglobal villageB of humans, linked bythe mass media; economists were beginning to arguefor increasing globalization of trade and commerce;and, in the earth sciences, the paradigm of platetectonics was having an enormous impact on ourunderstanding of earth history. Indeed, the globalreach of seismic stratigraphy was one of its mostpersuasive features.

In a companion paper Miall and Miall, in prepa-.ration , we examine social factors shaping the devel-

opment, dissemination, and initial validation of seis-mic stratigraphy and the social organizations in whichthese occurred. The main objective of the presentpaper is to examine the global-eustasy model fromthe period of the initial enthusiastic reception ofseismic stratigraphy by the scientific community inthe late 1970s, to a period of increasing doubt aboutthe global-eustasy model that extends to the presentday. Our main purpose is to show that beyond theundoubted new facts and invigorating new ideas thatseismic stratigraphy brought to geology, the popular-ity of the new paradigm within the geological com-munity owed much to human, AsocialB factors. Wedraw extensively on Kuhns work regarding the de-velopment and acceptance of new paradigms in sci-

( )A.D. Miall, C.E. MiallrEarth-Science Reiews 54 2001 321348 323

ence. The paradigm model has been usefully appliedto the study of several areas of research in the earthsciences, including the development of turbidite con-

.cepts Walker, 1973 , methods of grain-size analysis .in sedimentology Law, 1980 , the acceptance of

plate-tectonic theories about the earths crust .Stewart, 1986 , and the controversy surrounding the

.abiogenic origins of oil Cole, 1996 .Further, we argue here that because of controver-

sies surrounding ideas about the origins of se-quences, sequence stratigraphy has now evolved intotwo distinct, competing paradigms. During the 1980s,a series of anomalies and conceptual problems aboutthe global-eustasy model emerged. We argue thatthese have not been fully addressed by proponents ofthe global-eustasy model, many of whom continue touse this model as a central theme in their strati-graphic work, despite a growing controversy thatsurrounds it. We do not examine, or critique, theconcepts of sequence architecture that, as notedabove, were first clearly formulated in the late 1980s.As part of our analysis, we examine the nature ofstratigraphic data and its importance in the processof validating Vails new ideas. The paper is offered

.as a contribution to the study of Dotts 1998 ques-tion: AWhat is unique about geological reasoning?BWe also attempt to alert geologists to the bias thatpreconceptions and group processes can bring toobservations and interpretations in the geologicalrecord.

2. Data and argument in geology

The traditional view of science, as described by .the tenets of analytical philosophy Frodeman, 1995 ,

is that science proceeds from careful, objective ob-servation and replication of data, to hypotheses andtheory through the workings of the scientific method.

.According to Popper 1959 , scientists engage in thepractice of proposing and falsifying testable hypothe-ses, based on dispassionate experimentation. There issome basis for arguing that this is what actuallyhappens in the so-called AhardB sciences, whereideas may be tested by carefully designed experi-ments. Even here, however, human factors can inter-vene, as the controversy surrounding Acold fusionB afew years ago attests. The proponents of a new

model of cold fusion were able to convince them-selves that their experiments had provided evidencefor it, but attempts at replication by other observersdemonstrated the falsity of the original claims, and

the new hypothesis was quickly discarded Peat,.1989 .

It is not so simple in geology, where we areattempting to understand a past that cannot be repli-

.cated by experiment. As Dott 1998 pointed out,there is much that we can replicate, such as thechemistry of mineral formation, or flume experi-ments to model bedform generation, but we cannotrecreate the past in its complex entirety. The newscience of numerical modeling is providing us withpowerful new techniques for simulating complexprocesses, such as stratigraphic accumulation underconditions of varying rates of basin subsidence andclimate change, but these are just simple models, notcomplete replication experiments, so geologists have

.to search for what Frodeman 1995 called Aexplana-tions that work.B Much of geological practice consti-

.tutes what Dott 1998 referred to as synthetic sci-ence.

Our views of Awhat worksB have changed dramat-ically as the science has evolved. Consider the sci-ence of stratigraphy, for example. Until the 1950s,stratigraphic practice consisted primarily of what wenow term lithostratigraphy, the mapping, correlatingand naming of formations based on their lithologicsimilarity and their fossil content. In the 1960s, therevolution in process sedimentology led to the emer-gence of a new science, facies analysis, and a focuson what came to be termed AautogenicB processes,such as the meandering of a river channel or theprogradation of a delta. Most stratigraphic complex-ity was interpreted in facies terms, and the sciencewitnessed an explosion of research on processre-sponse models, otherwise termed facies models. Therevolution in seismic stratigraphy in the late 1970schanged the face of stratigraphy yet again, with anew focus on large-scale basin architecture and re-gional and global basinal controls.

In the course of 20 years, therefore, the kinds ofdata geologists looked for in the rocks, and theAexplanations that workedB in explaining them, un-derwent two wholesale changes. The rocks did notchange, but the AobjectiveB facts that geologistsextracted from them did. This attests to our improved


understanding of our own subject, but the details ofthe evolution of this science, as with any other, arealso influenced by human factors. The fact thatscientific journals include ACommentaryB or ADis-cussionB sections attests to the fact that apparentlydispassionate observation can, nonetheless, lead todifferent interpretations and to controversy. Scien-tists accept this, while they remain reluctant to ac-cept that human factors play an important role inscientific development. How important is the reputa-tion of the scientist in furthering a new idea? Howimportant is AfashionB? Those of us who were activein the 1970s may remember a time when turbiditemodels were first popularized, and all thinly bedded

lithic arenites tended to be reinterpreted as turbidites,and later, when the new sabkha model became theexplanation for all evaporites.

Social scientists have helped to show that science,like other human endeavors, is a human activity,subject to the same social influences as non-scien-

tific endeavors Kuhn, 1962, 1996; Cole, 1992;. .Barnes et al., 1996 . Mulkay 1979 , for example,

has concluded that physical reality constrains, butdoes not uniquely determine, the conclusions of sci-entists. Much depends on the preconceptions of theinvestigator. Further, the evolution of a body of ideasdepends on the social and work conditions in which

the practitioners find themselves e.g., Barnes et al.,

Fig. 1. The hermeneutic circle, as applied to the global-eustasy model. Each of the four stages in the circle is illustrated in lower-case .lettering. AGuiding principlesB is the term used by Vail 1992 . The recognition of anomalies should lead to AfalsificationB and the

emergence of new guiding principles, but we argue that this process has not taken place during the development of the global-eustasymodel.


.1996 . Particular data sets and models are not neces-sarily tidbits of universal truth to be teased from theether by dispassionate scientists, but are very much aproduct of the social conditions within which thescientific work is carried out. Even primary scientificobservations are not value-free, but have a context.This context may include an array of hypotheses;these may be made using special observation meth-ods, and these methods may reflect assumptions or

simplifications of their own Barnes et al., 1996, p.. .2 . Barnes et al. 1996 have made a distinction

between a simple AobservationB and that which ismade and reported as part of an hypothesis-testingexercise. The latter they have termed an obserationreport. They have suggested that Aour thoughts couldinfluence our perceptions as well as our perceptionscould influence our thoughts.B Indeed, it has beenargued that there really is no such thing as a ApureBobservation. There may be many assumption sets ina given field. For example, as noted above, the samestratigraphic section may have been described in atleast three different ways since the 1950s. As Kuhn .1962, p. 129 has noted, Aone and the same opera-tion, when it attaches to nature through a differentparadigm, can become an index to a quite differentaspect of natures regularity.B This is not intended asa criticism or negation of the scientific method, butas an attempt to throw light on the very humanprocesses involved in the action of Adoing science,Bso that we might understand it better.

Such an analysis of the very foundations of thescientific method, which points to and incorporatesthe various contextual characteristics of the work, isa dramatically different way of viewing science thanthe analytical approach. Based on the philosophicalideas of Heidegger, this approach is termedhermeneutics, and the back-and-forth thought pro-cesses between observation and interpretation are

conceptualized as the hermeneutic circle Heidegger,1927, 1962; Frodeman, 1995; Barnes et al., 1996;

. .see Fig. 1a . Frodeman 1995 , in a discussion of thephilosophy of science, noted that Heidegger identi-fied three types of prejudgement or forestructuresthat scientists bring to each situation: preconcep-tions; ideas or presumed goals, and a sense of Awhat

.will count as an answerB Frodeman, 1995, p. 964 ;and the particular tools and methodologies of theresearch. Hermeneutics argues that our original goals

and assumptions result in certain facts being discov-ered rather than others, which in turn lead to new

.avenues of research and sets of facts. Rudwick 1996argued that, beyond the basic data directly measur-able by technological means, such as the density ofpyrite, or the drilling depth of the Toronto Forma-tion, Athere are no theory-free facts in geology.BWhile extreme, this viewpoint provides a usefulcounterweight to the power of a newly popular the-ory.

The main purpose of this paper is to attempt toshow how Ahuman factorsB influenced, and stillinfluence, the progress of a single geological model,that of global-eustasy and its implications for se-quence stratigraphy.

3. The hermeneutic circle and the emergence ofsequence stratigraphy

Sequence concepts, as first developed by Sloss et . .al. 1949 and Sloss 1963 , constituted an entirely

new methodology for documenting and interpretingthe stratigraphic record, but had not become theprevailing stratigraphic methodology when ap-proaches based on seismic stratigraphy were firstintroduced in the mid 1970s. The new work thatemerged from Exxon in the 1970s brought aboutwhat Kuhn would term a AcrisisB in sedimentarygeology; that is, the introduction of a new methodol-ogy and a body of ideas that could not be reconciled

.with existing paradigms. As Kuhn 1996, p. 181 hasstated:

Crises need not be generated by the work of thecommunity that experiences them and that some-times undergoes revolution as a result. New in-struments like the electron microscope or newlaws like Maxwells may develop in one specialtyand their assimilation creates crises in another.

The development of high-quality reflection-seismic data and the new methods of stratigraphic

interpretation based on these data e.g., the focus onstratigraphic surfaces and terminations, and on

.three-dimensional architecture clearly constituted anew AinstrumentB in the specialty of petroleum geo-physics. These data then affected the field of conven-tional, largely university- and state-survey-based


stratigraphy. To this extent, Vails work created aAcrisis,B which his work was, in turn, designed toresolve. Indeed, the emergence of sequence stratigra-phy resulted in two additional challenges to what

.Kuhn 1962 would have referred to as the normalscience of stratigraphic interpretation:

.1 The proposal that the global cycle chart was asuperior standard of geologic time to that based onconventional chronostratigraphy. For example, Vail

.et al. 1977, p. 96 stated: AOne of the greatestpotential applications of the global cycle chart is itsuse as an instrument of geochronology.B Vail and

.Todd 1981, p. 217 stated, with regard to correla-tions in the North Sea Basin: Aseveral unconformi-ties cannot be dated precisely; in these cases theirages are based on our global cycle chart, with ageassignments based on the basis of a best fit with thedata.B They proceeded to revise biostratigraphic agesbased on the correlations suggested by their chart.

.2 The attribution of all changes in sea level totwo favored eustatic mechanismseustasy typically

.glacioeustasy for high-frequency sequences , andchanges in ocean-basin volumes Vail et al., 1977,

.pp. 9294 , and the assertion that other regionalprocesses, including tectonism and changes in sedi-ment supply, affected only the amplitude but not the

timing of sea-level changes Vail et al., 1991, p..619 .

Unlike that which occurred during the develop-ment and acceptance of plate tectonics Stewart,

.1986 , there was no AcrisisB in normal science, in the .sense originally intended by Kuhn 1962 . Other

geologists were not dissatisfied with the body ofAnormalB stratigraphic science. Conventional chrono-stratigraphic methods, based on biostratigraphy, ra-diometric dating, chemostratigraphy, etc., were, andstill remain, the primary means for determining geo-

.logic age e.g., Harland et al., 1990; Holland, 1998and, to the extent that there had been no widespreadsearch for global mechanisms for stratigraphic pro-cesses, most were comfortable with the prevailingviews regarding the complexity of geological pro-cesses. Nonetheless, it could have been expected thata geological community, alerted to the power and

influence of the new global tectonics plate tecton-.ics in the mid- to late 1970s, would view a new

global model with interest and excitement as we.document elsewhere: Miall and Miall, in preparation .

.Vail 1992 himself has attempted to describe hisown scientific procedures. In his memoir on theevolution of sequence stratigraphy, he discussed howresearch creativity could be optimized through theapplication of a set of procedures, which seeminglyreflected the hermeneutic approach. These includedwhat we could term forestructures, such as the estab-lishment of a thematic research program with clearlydefined goals for the overall research, and the defini-tion of concepts that would drive the thematic re-

searchdriving concepts. However, Vail 1992, p..84 argued that it was important to define driving

.concepts or Aguiding principlesB: Fig. 1 because itwould then be possible Ato acknowledge and nurtureideas that challenge, and may prove superior to the

existing driving concepts.B He also argued on the.same page that,

truly new, worthwhile ideas based on competingdriving concepts may not be accepted within theframework of thematic research. These competingdriving concepts are commonly ignored or putaside because of the priority of other work.

Despite this awareness that driving concepts inthematic research can direct attention away fromanomalous observations and hypotheses, and basedon his own recollections of his time at Exxon, Vailappears to have established, within his own seismicresearch group, a structure likely to yield the kind ofhermeneutic circle that did not allow for competingdriving concepts, as we argue in the remainder of

. .this section Fig. 1 . According to Vail 1992, p. 89 ,two organizing principles appeared to guide the re-search. First, a working environment was fostered inwhich problems were accurately defined and interre-lated through the establishment of a thematic re-search program informed by the driving conceptsmentioned above. These concepts directed thegroups attention to the problems to be solved, themethods to be used in obtaining solutions, and the

types of phenomena to be studied from Vail, 1992,.p. 87 :

What this driving concept showed was that seis-mic sections are a high-resolution tool for deter-mining chronostratigraphythe time lines inrocks. This was a AeurekaB at that time. We hadfound the tool and developed the methodology to


make regional chronostratigraphic correlations andto put stratigraphy into a geologic time frameworkfor mapping and the understanding of paleogeog-raphy . . . The fact that seismic reflections followtime lines is the second basic driving concept.

.As Sloss 1988, p. 1661 puts it,

the sequence concept was alive and well in aresearch facility of . . . Exxon. Here, Peter Vail anda cohort of preconditioned colleagues seized uponthe stratigraphic imagery made available by multi-channel, digitally recorded, and computer-mas-saged reflection seismography to establish thediscipline of seismic stratigraphy.

Second, an environment was created where bothteamwork and individual responsibility co-existed.

.As Vail 1992, p. 89 has observed, in his article onthe evolution of seismic stratigraphy and the globalsea-level curve,

as a group, we developed an overall plan. Wewould then try to identify the person who wasmost interested and knowledgeable for each task,and then endeavor to give each person a maxi-mum amount of responsibility for his or her pro-ject area . . . We tried to develop a situationwherein each researcher had a clear-cut area ofresponsibility, but we made sure it overlappedwith as many other areas as possible. This en-sured good communication, because each personwas vitally interested in what the others weredoing.

.What this also ensured, as Law and Lodge 1984might argue, was that each member of the group hadan interest and an investment in the success ofseismic stratigraphy and the global-eustasy model.The work of the Exxon team, as described by Vail, isa good example of the Asocially constructedB natureof group scientific work. As Clarke and Gerson .1990 have argued, scientific theories, findings and

facts are socially constructed although, of course,.based on observation and measurement . They note

that to solve research problems, scientists will makecommitments to specific theories and methods, toother scientists, to research sponsors, and to various

.organizations. Clarke and Gerson 1990, p. 184further directed attention to the importance of

A . . . structural conditions of work, and the concreteprocesses of actually collating different lines of evi-dence.B These, they argued, are critical to the emer-gence and maintenance of belief in the results of aparticular line of research in the face of what theyterm the Aburied uncertaintiesB of the research.

It must be emphasized that we are using theExxon work here to illustrate common themes inscientific research, not in order to criticize themethodology, but in order to help focus in on theimportance of group dynamics as the researchevolved and came under scrutiny by the wider geo-logical community.

4. Paradigms and exemplars

.As Kuhn 1962, p. 121 has noted, Agiven aparadigm, interpretation of data is central to theenterprise that explores it, but that interpretive enter-prise . . . can only articulate a paradigm, not correct

.it.B Kuhn 1962, pp. 2324 has further argued that,

the success of a paradigm . . . is at the start largelya promise of success discoverable in selected andstill incomplete examples. Normal science con-sists in the actualization of the promise, an actual-ization achieved by extending the knowledge ofthose facts that the paradigm displays as particu-larly revealing, by increasing the extent of thematch between those facts and the paradigmspredictions, and by further articulation of theparadigm itself.

The fundamental building blocks of new science,he goes on to state, are Asolved problems.B These hereferred to as exemplars, because they are used asexamples for teaching and in order to guide further

.research see also Barnes et al., 1996, pp. 101109 .In a historical science like geology, it is difficult

to arrive at AtruthB or AproofB by the normal processof experimentation and replication though not al-

.ways impossible, as demonstrated by Dott, 1998 . Itcould be argued that exemplars are Aexplanations

.that work;B what Frodeman 1995 has referred to asa type of understanding having narratie logic. Typ-ical examples of exemplars in this case are: AIn theNorth Sea, 15 of 17 unconformities identified on the


basis of well results and seismic data matchedExxons global sea-level curve.B P.R. Vail at a

meeting in Woods Hole Massachusetts in April as.reported by Kerr, 1980, p. 484 . And again: AWe can

w xcorrelate ten of them unconformities perfectly withthe Vail curve, five correlate pretty well, and the

others still have a few problems.B Tom Loutit andJames Kennett commenting on their work in New

.Zealand as reported by Kerr, 1980, p. 485 . In afollow-up report on the Vail method 4 years later,

.Kerr 1984 reported several more of these AX out ofYB comparisons, where Y is slightly greater than butnever equal to X. These observations began to formthe basis for empirical generalizations about sea-level

.events Fig. 1 .As we document elsewhere Miall and Miall, in

.preparation , the publication of exemplars of thistype helped to rapidly convince the geological frater-nity of the power and importance of the new global-eustasy model. This, despite the lack of supporting

.documentation in Vail et al. 1977 , including, forexample, interpreted seismic lines or biostratigraphi-cally documented subsurface sections. Missing,therefore, are the AobservationsB that are supposedly

so critical to the scientific method Barnes et al.,. .1996 . At best, we have what Barnes et al. 1996

might have called Aa report that such observationsexist.B Given the lack of published documentation,there was no opportunity for outsiders to influencethe workings of the hermeneutic circle on the devel-opment of the global-eustasy model.

In their discussion of the development of scien- .tific knowledge in general, Barnes et al. 1996

discussed how an important experiment in physics .determination of the charge of the electron wasgradually refined by repeated experimentation, docu-mented by extensive laboratory note-taking, and howthe results of a rival set of experiments that gener-ated different results were eventually discarded be-cause the results were gradually shown to be incon-sistent with the results of new work. Similarly, at

.some point, Vail et al. 1977 decided that a givenset of sea-level events represented global-eustaticsignals, and the first global cycle chart was theresult. The global-eustasy hypothesis seems to havebegun to work a powerful influence on the selectionand collation of additional data. But how were theseevents selected? How did Vail know that these were

the ArightB ones? How were other events discardedas the result of Alocal tectonics,B or as having beenAoffsetB by biostratigraphic imprecision? When con-sideration is given then to the Asolved problemsB orexemplars of the global cycle chart, the methodologyof documentation and cross-checking and virtuallyall of the primary data are missing. For example, thereaders attention is directed to the only published

diagram in AAPG Memoir 26 Vail et al., 1977, Fig..5 on p. 90 that illustrates the synthesis of individual

cycle charts into the global average. There are sev-eral events appearing in the average curve that arenot well represented in the individual charts, and

vice versa see also discussion of this subject in.Miall, 1991, 1997, pp. 309311 . These points have

not been explained by Vail et al.In his discussion of the consensus model of sci-

.ence, Kuhn 1962 has observed that the real func-tion of experiment is not the testing of theories. Heshowed that commonly, theories are accepted beforethere is significant empirical evidence to supportthem. Results that confirm already accepted theoriesare paid attention to, while disconfirming results areignored. Knowing what results should be expectedfrom research, scientists may be able to devise tech-

niques that obtain them Kuhn, 1982; Cole, 1992, pp..78 . The exercise of correlating new stratigraphic

sections to the global cycle chart entails the dangersof self-fulfilling prophecy. As noted by Kuhn 1962,

.pp. 80, 84 :

The bulk of scientific practice is thus a complexand consuming mopping-up operation that consol-idates the ground made available by the mostrecent theoretical breakthrough and that providesessential preparation for the breakthrough tofollow. In such mopping-up operations, measure-ment has its overwhelmingly most common scien-tific function . . . Often scientists cannot get num-bers that compare well with theory until theyknow what numbers they should be making natureyield.

The lack of published experimental documenta-tion and detailed analysis in support of the global-eustasy model makes it difficult for scientists ingeneral to evaluate the importance of these humanprocesses that Kuhn described.


Why did geologists eagerly embrace the global-eustasy model despite the lack of published data thatthey could see for themselves? We suggest that it

.was 1 because of the current willingness to acceptglobal explanations of earth processes in light of the

.new plate-tectonics paradigm; 2 the ideas offered autilitarian applicationthe global correlation tem-plateof apparently considerable potential use in

.the exploration business; 3 because of the assumedauthority of corporate geophysics, or what we term

.Athe Exxon FactorB Miall and Miall, in preparation ; .and 4 because working petroleum geologists had

no investment in the complex, confusing andAacademicB science of conventional chronostratigra-

.phy. Dott 1992 suggested an additional factor, theAinnate psychological appeal of order and simplic-ityB of a pattern of cyclicity based on Milankovitchperiodicity.

Use of the global cycle chart appeared to offer asimple global solution to the problem of stratigraphiccorrelation so, perhaps, it was not surprising thatworking geologists eagerly adopted it. Indeed, thenon-availability of a data base, with all the messi-ness, incompleteness and inconsistency such as nor-mally characterizes stratigraphic successions fromdiverse areas characterized by different tectonic his-tories, undoubtedly made it easier to accept the

.global cycle chart as is. To cite Laws 1980, p. 13summation of Mannheims philosophy, the applica-tion of the global cycle chart displayed

the attributes of Mannheims natural law style ofthoughtthey are atomistic, generally quantita-tive, emphasize the routine nature of scientificpractice, seek or utilize general laws, stress conti-nuity, and are in general reductionist.

The techniques for identifying sea-level events in .seismic records and later stratigraphic sections, and

equating them to existing events in the global cyclechart, became a routine operation ideally suited topetroleum-exploration work. The alternative, holisticapproach to stratigraphy is the traditional one ofcorrelation by biostratigraphy and the erection ofstratotypes with no built-in assumptions of global-events. It was this approach that Vails chart seemeddestined to replace. While Vail et al. have repeatedlyreaffirmed the importance of biostratigraphic dating

and correlation in the testing of the global cyclechart, the emphasis in much of the work of thisgroup is on the superiority of the global cycle chartas a method of correlation and stratigraphic standard-

ization Vail and Todd, 1981, p. 230; Vail et al.,1984, p. 143; Baum and Vail, 1988, p. 322; Vail etal., 1991, pp. 622, 659; see summary in Miall, 1997,

.p. 282 .In the early 1980s, however, problems with the

global-eustasy model began to emerge as alternativeideas about sequence generation began to appear,and doubts emerged about the accuracy and preci-sion of chronostratigraphic methods available to testglobal correlations. We refer to these developments

in the succeeding paragraphs they are discussed in. detail in Miall, 1997 . Carter et al. 1991, pp. 42 and

.60 pointed out that sequence stratigraphy embracestwo quite distinct concepts:

Two different conceptual models underlie the ap-plication of sequence stratigraphy by Vail et al.:

w xone model relates to presumed global sea-levelbehaviour through time; the other model relates tothe stratigraphic record produced during a singlesea-level cycle. Though the two models are inter-related they are logically distinct, and we believethat it is important to test them separately . . . Ourstudies lead us to have considerable confidencein the correctness and power of the Exxon se-quence-stratigraphic model as applied to sea-levelcontrolled, cyclothemic sequences . . . At the sametime, we suspect that the Exxon Global sea-levelcurve, in general, represents a patchwork throughtime of many different local relative sea-levelcurves.

We suggest that the concerns about the global-eustasy model and the global sea-level curve haveled to the emergence of two competing paradigms,which we now discuss.

5. Two competing paradigms

In his essay on the structure of scientific revolu- .tions, Kuhn 1962, p. 65 has noted that anomalies

.data that do not AfitB appear only against thebackground provided by a paradigm. In this regard,he has argued that Aif an anomaly is to evoke a


crisis, it must usually be more than just ananomaly . . . Sometimes an anomaly will clearly callinto question explicit and fundamental generaliza-

.tions of the paradigmB Kuhn, 1962, p. 82 . As Kuhn .1962, p. 76 has also observed, however, Aphiloso-phers of science have repeatedly demonstrated thatmore than one theoretical construction can always beplaced upon a given collection of data.B As wedocument below, views about the global-eustasymodel have evolved into a broad spectrum of ap-proaches, and it is useful to consider them as two,coexisting, end-member paradigmsthe global-eus-

tasy paradigm the original concepts set out in Vail.et al., 1977 , and what we label the complexity

paradigm. Sequence stratigraphy, itself, includingstudies of sequence architecture, could be said to

.correspond now to Guttings 1984 definition of asupertheory: a Awide-ranging set of fundamental be-liefs about the nature of some domain of reality andabout the proper methods by which to study thatdomain.B

5.1. The global-eustasy paradigm

Eustatic sea-level changes are global by defini-tion, but we employ the word AglobalB here in thesense of Aall-encompassingB or Auniversal,B becausethis is the sense in which the model of eustaticsea-level change is employed by Vail et al. Thisgroup of workers attributes all sea-level change toeustatic processes, including changes in ocean basinvolumes for the low frequency cycles, andglacioeustasy for those of high frequency, although itwas conceded that there is no convincing evidence

for glaciation prior to the Oligocene Vail et al.,.1977, pp. 9294 . Two quotes from Vails first

major publication states the case succinctly:One of the greatest potential applications of theglobal cycle chart is its use as an instrument ofgeochronology. Global cycles are geochronologicunits defined by a single criterionthe globalchange in the relative position of sea level throughtime. Determination of these cycles is dependenton a synthesis of data from many branches ofgeology. As seen on the Phanerozoic chart . . . theboundaries of the global cycles in several cases donot match the standard epoch and period bound-

aries, but several of the standard boundaries havebeen placed arbitrarily and remain controversial.Using global cycles with their natural and signifi-cant boundaries, an international system ofgeochronology can be developed on a rationalbasis. If geologists combine their efforts to pre-pare more accurate charts of regional cycles, anduse them to improve the global chart, it canbecome a more accurate and meaningful standard

.for Phanerozoic time Vail et al., 1977, p. 96 .

Glaciation and deglaciation are the only well-un-derstood causal mechanisms that occur at the

relatively rapid rate of third-order cycles Vail et.al., 1977, p. 94 .

As noted above, Vail et al. thought that they haddeveloped an entirely new standard of geologic timeand a new basis for geologic correlation, superior tothat based on conventional chronostratigraphic databiostratigraphy, magnetostratigraphy, radiometric

. .dating see summary in Miall, 1997, pp. 282284 .

5.2. The complexity paradigm

This contrasting paradigm represents a body ofideas focusing on the hypothesis that sea-level changeis affected by multiple processes operating simulta-neously at different rates and over different ranges oftime and space, possibly including eustatic sea-levelchange. No simple signal can therefore result, andthe interpretation of each event in the stratigraphicrecord can only follow very careful, detailed localwork, followed by meticulous regional and globalcorrelations. An additional, and important argumentis that because of limitations in our techniques fortesting the reality of global-eustasy, no conclusionscan yet be drawn regarding the vast majority of thesupposed eustatic events that have been proposed .e.g., Miall, 1997; Dewey and Pitman, 1998 .

Underlying both paradigms are the principles andmethods encompassed by the sequence-architecture

.model first introduced by Vail et al. 1977 , and .further elaborated by Haq et al. 1988 , Posamentier

. .and Vail 1988 , Posamentier et al. 1988 , and Van .Wagoner et al. 1990 , although details of definition

and interpretation of the model remain to be re-solved. The global-eustasy model and its global cy-cle chart could be said to represent the paradigm


against which anomalies have arisen, and which havecoalesced around the complexity paradigm.

As others have already argued, geology is charac-terized by complexity. The concept of the multiple

working hypothesis originated with geologists Gil-.bert, 1886; Chamberlin, 1897 and, as noted by Law

.1980, p. 16 , it is not uncommon for geologists tosimultaneously consider contrasting or even conflict-ing methods and hypotheses, a condition he refers toas conceptual pluralism. Therefore, the growth oftwo conflicting views about global-eustasy and their

mutual survival for more than 15 years since the

first articles critical of the model appeared in the.early 1980s should not surprise historians or sociol-

. ogists of science or geologists . As Kuhn 1962, p..149 has observed, Athe proponents of competing

paradigms practice their trades in differentworlds . . . practicing in different worlds, the twogroups of scientists see different things when theylook from the same point of view in the samedirection.B

.Kuhn 1996, pp. 181187 has also clarified theconcept of the scientific paradigm by showing how itis constructed of a set of integrated parts. First, a

Table 1The components of the two paradigmsa

Component Global-eustasy paradigm Complexity paradigm . .Symbolic generalizations or 1 Seismic reflections are chronostrati- 1 There are few global AsurfacesB beyond those caused

laws graphic surfacesthey define time by very rare catastrophic events e.g., terminal Creta- . .2 Tectonism may AenhanceB an uncon- ceous event

.formity but does not affect its timing 2 Seismic reflections, when defined using the mostadvanced processing techniques, do not yield simpleglobally correlatable surfaces .3 Tectonism, eustasy, and varying sediment supplyintegrate to generate highly diachronous sequenceboundaries

. .Commitments or beliefs in par- 1 Sequences are chronostratigraphic units 1 No single technique can be used to precisely define .ticular models 2 Global eustasy is the dominant mecha- geological time

.nisms for driving relative sea-level change 2 There are multiple causes of sea-level change, ofvarying frequency and areal extent

. .Values 1 The conventional time scale is arbitrary 1 Only by defining boundaries withinin defining boundaries within continuous continuous successions can we be sure wesuccessions are defining a record of continuous time . .2 Sequence chronostratigraphy is superior 2 Only a compilation and reconciliation of all age datato conventional methods for dating and can begin to approach an accurate and precise time scalecorrelation

.Values imply prediction Global eustasy permits worldwide correla- 1 Where precise chronostratigraphic methods can betion used they will generally show that precisely dated

sea-level events in different parts of the world do NOTcorrelate with each other .2 Given the imprecision of chronostratigraphic meth-ods, most data strings will be shown to be capable ofcorrelation with the global cycle chart, even syntheticsections compiled from tables of random numbers

.Exemplars Asolved problemsB 1 X out of Y unconformities in a given Sequences in a given succession commonly can be.or Aways of seeingB data set match the global cycle chart correlated to localrregional tectonicrclimaticrautogenic

.2 A previously known series of events can processesbe shown to AmatchB or Acorrelate withB

the chart e.g., Alberta Basin molasse pulses,.Pyrenean thrust-fault pulses

a .The components are those defined by Kuhn 1996, pp. 181187 .


paradigm incorporates a system of symbolic general-izations or laws. Second, it involves commitments orbeliefs in particular models. Third, there are ideasabout values, such as acceptable levels of quantita-tive rigor. Fourthly, predictions are implied by theparadigm. Fifth, as already discussed, there are theexemplars, or solved problemsAways of seeing.BThese components of the two paradigms under con-sideration here are shown in Table 1. In Section 6,we discuss how the assumptions of these paradigmsguide the observations geologists make in Areality,Band impact on the conclusions they draw and theanomalies they do or do not confront.

6. Defining and deconstructing global-eustasy andcomplexity texts

In this section, we consider how the use of termi-nology and language in sequence stratigraphy mayinfluence perceptions and analysis, although we donot deny the existence of the objective ArealB worldnor the role it may have in the production of scien-tific knowledge. Rather, we attempt to direct atten-tion to how socially derived meanings and processescan shape how science, and in this instance, se-quence stratigraphy is done.

Social theorists examining how humans interact inmeaningful ways have argued that Adefinitions ofsituationsB guide the process. Through the process ofsocialization and the learning of language, humansattribute meaning to the situations in which they findthemselves whether these be processes of eating,socializing, or doing science. Specifically,

the crucial fact about a definition of a situation isthat it is cognitiveit is our idea of our locationin social time and space that constrains the waywe act. When we have a definition of a situation,we cognitively configure acts, objects, and othersin a way that makes sense to us as a basis for

.acting Hewitt, 1997, p. 127 .These theorists have as a central focus, the subjec-

tive standpoint of individual actors. AUnlike a moreobjectivist approach, which views the social world asa reality that exists independently of any individualsperception of it, phenomenology sees that reality as

constituted by our view of itB Hewitt, 1997, pp..1516 .

Most interaction among humans takes place withinroutine situations with well-established definitions ofthe situation or subjective viewpoints that guide be-

.haviour. According to McHugh 1968 , there arethree fundamental assumptions that influence routine

.interactions: 1 that the assumptions we hold about .a situation are valid; 2 that others in the situation

.share our definition of the situation; and 3 that aslong as our definition of the situation works, it willnot be questioned or challenged. These theorists,therefore, emphasize the cognitive foundations ofhuman conduct, manifested in language, and stressthat what people know about a situation and what

.they do are interdependent cf. Hewitt, 1997 .We suggest that through the application of these

principles to the study of science as a human activ-ity, we can understand how different groups of scien-tists, with different convictions or definitions of thesituation, could emphasize or de-emphasize particu-lar types of data and hypotheses in favour of thedesired goals of the research. In the global-eustasymodel, for example, published supporting data aresparse, and the necessary supporting hypotheses arelargely AinternalB or Aself-referential.B Accordingly,the language of the global-eustasy school includes anumber of key interpretive phrases, which serve toturn research results toward the desired model, todefine the situation confronting the geologist in aroutinized way that directs his or her analytic be-haviour. Notably, these terms are not used by, orhold different meaning for, the nonmembers of thisgroup, for example, those adhering to the complexityparadigm, who have their own set of definitions ofthe situation. Some of these terms that hold specialmeaning for the global-eustasy school, but are usedselectively or not at all by other sequence stratigra-phers, are presented in Table 2. They express two ofthe three elements of the hermeneutic circlethepreconceptions and the presumed goals of the re-search.

In order to further understand and illustrate theconsequences of the use of particular terms in thepractice of geology, it is useful to draw on a form ofliterary criticism called Deconstructionism. As pro-posed by Jacques Derrida, who founded the ap-proach, deconstructionism takes apart the logic oflanguage in which authors make their arguments.

.According to Denzin 1992, p. 32 , Ait is a process


Table 2The language of the global eustasy schoolTerm Application in the context of the global Usage by followers of the complexity paradigm

eustasy model

Eustasy The universal control of sequence bound- A hypothesis for which the evidence is sparse andaries questionable

Global cycle chart Universal stratigraphic template, applicable A compilation of local and regional events incautiouslyworldwide labeled as AglobalB

Glacioeustasy The mechanism for all high-frequency cy- Operated only during Earths major glacial periods, andcles must always be tested and calibrated against local and

regional tectonic and other mechanismsSequence chronozone Sequence interpreted as a primary chronos- A hypothetical concept of no practical use c.f. Hedbergs

. .tratigraphic unit 1976 chronostratigraphic unitsSechron Same as above Not used1st5th-order cycles The sequence hierarchy, based on an as- An arbitrary classification not reflecting the documented

sumed grouping of cycle frequencies variability in stratigraphic unit thickness or durationTectonically enhanced uncon- An unconformity of eustatic origin, en- Term not usedformity hanced by upliftTectonic overprint Eustatic unconformity modified by local Tectonic overprint on any other process, such as climate

tectonics change, variations in sediment supply, or eustasyLocal tectonics The reason why a sea-level event in the Just that: local tectonics, with no specific generalized

global cycle chart is absent from a particu- meaninglar stratigraphic section

Number of sequences in a sec- Used as an attribute of correlation An incidental result of local processes with no globaltion significancePattern matching Technique of recognizing similar succes- A technique of questionable value given the several

sions of stacked sequences to those in the mechanisms for sequence generation that could beGCC as an attribute of correlation with the overprinted in any given successionGCC

Key words and phrases that hold special meaning for the members of this school, but are used selectively or not at all by other sequencestratigraphers.GCC: global cycle chart.

which explores how a text is constructed and givenmeaning by its author or producer.B It rejects theassumption that texts have logical meanings andargues for Ademystifying texts instead of deciphering

.themB Vogt, 1999, p. 74 . To use an example,objectivist studies of history A . . . assume the facticityof objects of historical analysis as constituted prior

.to the observers study of themB Hall, 1990, p. 26 .Deconstructionists challenge,

the artificial coherence of historical accounts, byshowing how to locate transcendent staging de-vices in historical discourses. Situated outsidehistory, such devices render historical accountsplausible to readers by providing history withcontinuity and discourse with meaning thema-tized by aboutness. A history of Nixons Wa-tergate crisis, for example, can only be narrated

by telescoping events into a coherent story Cohen,

.1986, pp. 7476 . In this light, any notion that thehistorical object is simply out there, waiting forthe historian to discover and describe it, seems a

.self-serving conceit. Hall, 1990, p. 26 . .As Hall 1990, p. 35 further observed, Ain an age

when deconstructionists are busy assaulting texts asinternally ordered assemblages, historical narrativehas become suspect as a special kind of storytelling.B

We do not subscribe to the extreme relativism ofpostmodernism and deconstructionism, which arecharacterized by Aan extreme or complete skepticismof, or disbelief in, the authenticity of human knowl-

.edge and practiceB Dawson and Prus, 1995, p. 107 .Rather, we use the tenets of deconstructionism as aheuristic device to illustrate the role of language inscientific interpretations of the ArealB world.

We can illustrate how phenomenological and de-constructionist approaches inform the scientific pro-


cess of stratigraphic interpretation with a simple .diagram Fig. 2 . Two interpretations of a strati-

graphic AtextB are shown. The text consists of a suiteof seven data points in a timespace universe, suchas sea-level lowstands within a tectonostratigraphicprovince, dated according to the best available meth-ods. In Fig. 2A, we see the application of theglobal-eustasy AlanguageB to the text by the firstgroup, those who accept the primacy of the global-eustasy model. One of the data points, chosen, per-haps, because it represents an important type section,is interpreted as representing a global sea-level event,and, based on the assumptions of the global-eustasylanguage, all nearby points are referred to it. This issuggested by the arrows, which imply correlation,not movement of the data points, although implicit inthe global-eustasy model is the idea that an eustaticevent, as observed at a given location, can serve todefine its age everywhere. Such statements as A13out of 15 events in my data correlate with sequence

boundaries in the global cycle chartB e.g., see Kerr,.1980, 1984 are essentially applications of the lan-

guage illustrated in Fig. 2A. A powerful and entirely

Fig. 2. Two approaches to the correlation of the same set ofstratigraphic events. The seven events are shown in the same

.timespace relationship in both diagrams. A A sequence bound- .ary in one section circled dot is assumed to represent a eustatic

event, and according to the global-eustasy model, all stratigraphi- .cally nearby events are interpreted to correlate with it. B Age,

with error bar, is determined independently for each location. Thetwo parallel lines indicate one standard deviation around the firstdated event, at the left. The degree of correlation of each eventwith the first event is then assessed on a statistical basis.

self-referential hermeneutic circle is at work in theconstruction of this type of model. Anomalies do notarise because the research is directed toward defininghow observations fit.

By contrast, Fig. 2B shows the same data inter-preted employing the language of the second group.Each data point is accompanied by a bar showing theAstandard errorB associated with the dating method,and the AlanguageB of interpretation also includes anerror band extending the standard error of the firstdata point across all seven points. The users of thedata are then free to decide, based on their assess-ment of error, whether or not some of the data pointsfall outside the band they would consider to definean acceptable level of correlation with the first point.In this way, the light each new point throws on thequestion of sea-level change in this area can beassessed quantitatively, point by point, based onassessments of the various generation mechanisms.

Here, the definition of the situation or hermeneutic.circle allows for greater consideration of anomaly in

observations. Consequently, no quick conclusionsare drawn by geologists using this approach andanomalies are more likely to be acknowledged. As

.Thomas 1931 has observed, if you define a situa-tion as real, it is real in its consequences. For theprotagonists of the two paradigms, each interpreta-tion of the stratigraphic texts is ArealB within theterms of their paradigm. For the adherents of theglobal-eustasy model, the AconsequenceB is that dis-cussion of the potential error in, or revision of, theage of a given sequence boundary is typically notundertaken, whereas error and revision are an inte-gral part of the complexity paradigm.

In Section 7, we discuss the interaction and de-gree of coexistence that the two paradigms exhibit,through an examination of the patterns of cross-cita-tion that exist between them.

7. Invisible colleges and the advancement ofknowledge

The study of citation patterns has long been anissue in social research on scientific knowledge and

.how it advances. Price 1961, 1963 , for example,observed that some disciplines continually referencetheir classic or foundational works. Other disciplines


cite only the most recent papers in a rapidly advanc-ing research front. Citations also reveal clusters ofinterrelated researchers and underlying social net-works, what Price referred to as Ainvisible colleges.B

.According to Crane 1972, pp. 13839 ,

analysis of the social organization of researchareas in science has shown that social circles haveinvisible colleges that help to unify areas andprovide coherence and direction to their fields.These central figures and some of their associatesare closely linked by direct ties and develop akind of solidarity that is useful in building moraleand maintaining motivation among members.

However, she goes on to argue, in order forscience to advance, Athe exchange of ideas is impor-tant in generating new lines of inquiry and in produc-ing some integration of the findings from diverse

.areasB Crane, 1972, p. 114 . Indeed, she concludes,scientific communities may become completely sub-jective and dogmatic if they are unable to assimilateknowledge from other research areas. We have al-ready discussed how the two paradigms under re-view have been constructed to yield different kindsof observations. We shall now consider the extent towhich each addresses the anomalies arising from theresearch of the other, and the consequences this mayhave for the advancement of scientific knowledge.

As a generalization, workers in the complexityparadigm cite the global-eustasy model as one ofseveral possible scenarios for geologic interpretation,but the reverse is typically not the case. Thorne .1992 , in a discussion of the assumptions underlyingsequence stratigraphy, and which is generally sup-portive of the methodology, noted that Athe strati-graphic literature since 1977, citing the Exxon sea-

level cycle charts, is prodigious An ARCO libraryauthor citation search of P. Vail contains 983 cita-

.tions between 1977 and 1988 . Various stratigraphictechniques have been used to test the applicability ofthe global sea-level cycle chart.B Adherents of theglobal-eustasy paradigm may cite opposing work,but rarely make the connection between that workand its implications for their preferred model. Evenpublications by basin analysts from other large andprestigious petroleum companies, such as British

.Petroleum e.g., Hubbard, 1988 have had little ef-fect on the use of the model by its adherents. For

.example, Vail et al. 1991 discussed Hubbardswork on the tectonic origins of sequence boundaries,but made no connection between these results andtheir implications for the generality of his global-eus-tasy model.

An opportunity to examine the pattern of citationused by each paradigm has been offered by theappearance of two recent special publications of the

.Society for Sedimentary Geology SEPM , both pub-lished in 1998. The first, a book entitled, Mesozoicand Cenozoic Sequence Stratigraphy of European

.Basins de Graciansky et al., 1998 of which PeterVail is a co-editor and contributing author, takeswhat may be described as the classic global-eustasyapproach to regional sequence studies, as clearlystated in the second paragraph of its Preface:

Sequence stratigraphy applies the inherent premisethat eustasy represents a global signal among thevariables that play a role in shaping depositionalsequences. This global signal plays an essentialrole in shaping depositional sequences laid downin response to changes in relative sea level. Be-cause of this global signal, bounding surfaces of

depositional sequences sequence boundaries at.their correlative conformities can be expected to

be synchronous between basins. To demonstratesuch synchroneity requires a very high strati-graphic resolution and a calibration of all strati-graphic disciplines.Although the book consists of a series of regional

studies, in which global correlation was not a statedobjective, the assumption that eustatic control maybe demonstrated by such studies is conveyed by thisquotation, also from the Preface:

The composite stratigraphic record of higher ordereustatic sequences shows a significant increase inthe number of sequences identified in the variousEuropean basins. Entries on the new charts in-clude a composite stratigraphic record of 221sequence boundaries in the Mesozoic and Ceno-zoic, compared to 119 sequences for the same

.interval identified by Haq et al. 1987, 1898 .Published at about the same time as the de Gra-

.ciansky et al. 1998 text is the book Paleogeo-graphic Eolution and Non-Glacial Eustasy, North-ern South America, compiled and edited by Pindell

.and Drake 1998 . Again, the Preface to this book


sets out the philosophical approach adopted by theeditors and contributing authors:

Our ability to isolate eustasy in most settings isseriously hindered by the fact that relative sea-level history in any location is multi-variable,comprising the net effects of such components aseustasy, local and regional tectonism, local andglobal climatic variability and variable sedimentsupply.

One of this books prime conclusions . . . is thatthe eustatic component of short-term relative sea-level changes during non-glacial times cannot beconfidently isolated in most settings. The inabilityto isolate short-term eustatic changes will pre-clude genetic correlation of short-term cycles andthe use of short-term cycles on cycle charts astime scales of predictors of reservoir horizons inmost basins.In Table 3, we list 26 papers that may be regarded

as the critical body of work exploring problems withthe global-eustasy paradigm. In a few cases, somerows in this table contain more than one paper. Thisindicates that the relevant ideas have appeared indifferent form in several places. Omitted from thislist are a few contemporary publications, such as

.Miller and Kent 1987 that, while making importantpoints, were not published in journals normally readby petroleum geologists, and were, therefore, notwidely noted or cited at the time of publication. Also

.not included is Pitman 1978 , an important earlywork that discussed rates of sea-level change, partic-ularly with respect to the rapidity of the sea-levelfalls implied by Vails first AsawtoothB curve, butdid not constitute a critique of the chart itself. Thelist begins with the work of Watts et al., whodeveloped the flexural model for extensional conti-nental margins. This model provides a powerfulalternative model for patterns of coastal onlap ofAsecond-orderB type and thus challenges the assump-tions of eustatic sea-level rise incorporated into theglobal cycle chart. Many of the authors listed inTable 3 also contributed to a series of Discussions of

.the Haq et al. 1987 paper that appeared in 1988,with Replies by the Exxon group Christie-Blick et

.al., 1988 .Table 4 documents the pattern of citation of these

critical articles by the contributing authors of the 45

.papers in the de Graciansky et al. 1998 volume,less the one article that is published only in abstractform. The results confirm the assertion that data andarguments, opposing the generality of the global-eus-tasy model generated by other specialists, have notbeen widely responded to by adherents of theglobal-eustasy paradigm. Based on this citation anal-ysis, the following observations may be made:

.1 Thirty-four of the 45 papers cite the global .cycle chart of Haq et al. 1987, 1988 , mostly by

showing how their sequences correlate to events inthat chart.

.2 Twenty-eight of the papers cite none of thearticles critical of the global-eustasy model, or anyother critical articles not on our list. None cite morethan five of those given in Table 3. Critical papers 2,3, 5, 6, 8, 10, 18, 19, 20, and 24 from Table 3 arenot cited at all. The Discussions and Replies of the

.1987 version of the chart Christie-Blick et al., 1988are not cited anywhere in this book.

.3 Most articles acknowledge the importance of .low-frequency second-order tectonism in generat-

ing accommodation for sequences, but only one citesthe early work in this area Paper row a1 in Table

.3 . A few papers acknowledge the potential impor- .tance of in-plane stress papers a4, 5, 11 .

.4 Tectonic AenhancementB or AoffsettingB ofeustatic sequence boundaries is a common theme.

.5 Several papers studied the influence of high-frequency tectonism in foreland basins and con-cluded that it is very local in effect and had no effecton the timing of sequence boundaries.

.6 A few papers cite the references discussing theproblems with the imprecision of chronostratigraphic

.methods papers a16, 21, 25 , but none of themattempt to answer or deal with these problems.

. .7 Only one paper that by Vecsei et al. showserror bars for the ages of sequence boundaries.

.8 The Anumber of sequencesB in a given timeinterval is regarded as a significant point of compari-son to the global cycle chart.

.9 Many papers claim that sequences must beglobal because they correlate with the global cyclechart, and use such correlations to extract or isolatethe eustatic component.

.10 A few papers note the circularity of correlat-ing with portions of the global cycle that weredefined in the general vicinity of their project area.


Table 3Key papers expressing doubt or opposition to parts or all of the global-eustasy modelNumber Authors Type Brief summary of main points

.1 Steckler and Watts 1978 , M Flexural model for onlap on extensional continental margins . .Watts 1981, 1989 , Watts et al. 1982

.2 Parkinson and Summerhayes 1985 M Integrating tectonism and eustasy does not generate cleareustatic signal

.3 Summerhayes 1986 S Data for GCC largely from Atlantic and Gulf margins, socorrelations there no proof of eustasy

.4 Cloetingh et al. 1985 , M First proposal for significance of in-plane stress as a mecha- .Cloetingh 1986 nism for RSLC

.5 Karner 1986 M Geophysical basis for in-plane stress .6 Miall 1986 S General critique of methodology of extracting sea-level history

from seismic data .7 Burton et al. 1987 , S Too many unknowns to permit extraction of eustatic signal

.Kendall and Lerche 1988 .8 Hubbard 1988 M Documentation of ages of sequence boundaries in several

major basins that do not correlate to GCC seismic data from a.ArivalB corporation: BP

.9 Blair and Bilodeau 1988 M Tectonic mechanisms for clastic-wedge generation .10 Algeo and Wilkinson 1988 S Cycle thicknesses are statistically not periodic and cannot be

used to argue for Milankovitch control. .11 Cloetingh 1988 M Detailed application of in-plane stress model to basinal

stratigraphy .12 Galloway 1989 M Sequences may reflect variations in sediment supply caused by

tectonic events in source area .13 Schlager 1989, 1991 M Some sequence boundaries are Adrowning unconformitiesB

.14 Embry 1990 M Methods for identifying tectonic influences in sequence gener-ation

.15 Christie-Blick et al. 1990 , M, S General critique of seismic methodology; emphasized impor- .Christie-Blick 1991 tance of tectonism

.16 Underhill 1991 M Tectonic origin for events in Moray Firth Basin that had beenused in definition of the GCC

.17 Miall 1991 C, M Chronostratigraphic imprecision of sequence boundary ages.Tectonic origin of sequences, critique of onlap model

.18 Aubry 1991, 1995 C Detailed tests of chronostratigraphic correlations of uncon-formities worldwide reveal no clear global signal

.19 Fortuin and de Smet 1991 M High rates of tectonism can occur on convergent plate margins .20 Sloss 1991 M Tectonism must be considered as a primary mechanisms for

sequence generation .21 Miall 1992 C GCC will correlate with any succession of events, including

synthetic sections compiled from random numbers .22 Gurnis 1992 M Dynamic topography: a mechanism for sequence generation

on cratons .23 Underhill and Partington 1993a,b M Tectonic origin for a major event in the central North Sea that

had been used in definition of the GCC .24 Drummond and Wilkinson 1993, 1996 S Cycle thickness reveals continuous distribution, therefore there

is no cycle hierarchy .25 Miall 1994 C The inherent imprecision of chronostratigraphic methods

global correlations cannot yet be reliably tested .26 Christie-Blick and Driscoll 1995 M Review of sequence-generating mechanisms, including tecton-

ism as a major factor

.Abbreviations: GCCsglobal cycle chart, RSLCs relative sea-level change. Types of critique column 3 : Csdiscusses chronostrati-graphic methods and the nature of detailed tests of the GCC, Msdiscusses alternative mechanisms for the generation of sea-level events,Ssgeneral critiques of methodology, circularity of correlations to GCC.


Table 4Citations of critical work of the complexity paradigm in SEPM Special Publication 60

aAuthor GCC cited Complexity citations CommentaryvHardenbol et al. none No discussion of potential error in correlationvJacquin and de Graciansky 1, 4, 14, 22, 23 Acknowledges tectonic influences but does not address the

implications of integrating tectonism and eustasyvJacquin and de Graciansky 11, 14, 23vDuval et al. nonevVecsei et al. 7, 13, 15, 17 SBs as drowning unconformities, SBs AoffsetB by tectonism.

Eustatic component isolated by correlation to GCC. The onlypaper to show SBs with error bars

vAbreu et al. none Cretaceous glacioeustasyvVandenberghe and none

HardenbolvNeal and Hardenbol none Paleogene sequences increased from 26 to 36vMichelsen et al. none SBs Abroadly comparableB to those in GCCvVandenberghe et al. nonevCatalano et al. none Tectonism enhances SBs but does not change their age

Vitale 9 Tectonically formed sequences correlate to GCC. High-frequencytectonism is only local

Flinch and Vail none High-frequency tectonism purely local and does not affect timingof SBs

vVakarcs et al. 4 Sequences are global because they correlate with GCCvGnacolini et al. none Sequences are global because they correlate with GCCvAbreu and Haddad 11, 25 Chronostratigraphic problems acknowledged but not dealt withvNeal et al. nonevGeel et al. 4, 11, 14 Applies Embrys criteria for tectonism but argues that this does

not exclude eustasy as a controlvLuterbacher none Correlations with GCC may be artifact, notes danger of circular

reasoningvPujalte et al. none Correlations with GCC may be because it was defined on data

from same part of EuropevHardenbol and Robaszynski none ANumber of sequencesB and Asequence stackingB criticalvGrafe and Wiedmann 13, 21, 26 References cited but points not dealt with

Floquet none Acknowledges importance of low-frequency tectonism but not itsimplications for eustatic correlations

vRobaszynski et al. nonePhilip none

vJacquin et al. noneRuffell and Wach none

vHoedemaeker none Uses French biostratigraphic system that does not always matchGCC. ANumber of sequencesB significant

vJacquin et al. 23de Graciansky et al. 23 Some successions of SBs cannot be correlated to GCC Awith

confidenceBStephen and Davies 14, 16, 23 Tectonism has not overprinted eustasy in North Sea, but Aun-

soundB to use such a Alocal datasetB to invoke global mechanismvLeinfelder and Wilson none

Gygi et al. nonevVan Buchem and Knox nonevHesselbo and Jenkyns 12, 13, 17, 23 References cited but points not dealt with

de Graciansky et al. 23vDumont 4, 11 Poor correlation to GCC may indicate GCC Adoes not apply

exactly to the European realm because global eustasy is not thesingle forcing mechanismB

vGianolla and Jacquin none Tectonic modification of eustatic events


.Table 4 continuedaAuthor GCC cited Complexity citations Commentary

vSkjold et al. 4, 11 Intraplate stress may have been importantvGoggin and Jacquin nonevCourel et al. nonevGaetanio et al. 14 Notes circularity of correlation to GCC that was mainly con-

structed from this project area. ATectonism versus eustasy: aneverlasting debateB

vGianolla et al. noneRuffer and Bechstadt none

.Abbreviations: GCCsglobal cycle chart of Haq et al. 1987, 1988 , SBssequence boundary.a The numbers in this column correspond to the numbered papers in Table 3.

By contrast, referencing by adherents of the com-plexity paradigm is much more inclusive of the otherparadigm. For example, the paper by Dewey and

. .Pitman 1998 in the Pindell and Drake 1998 vol-ume, which examines Asea-level changes: mecha-nisms, magnitudes and rates,B cites the followingarticles from Table 3: papers 1, 2, 4, 5, 6, 17, 21, and22. All of these articles are assessed in some detail inthe text of the paper. This paper also cites the majorworks by Vail et al., beginning with the classic Vail

.et al. 1977 volume. .The paper by Dewey and Pitman 1998 con-

cludes that there are multiple causes of sea-levelchange, most of them local to regional in scope. Inthis book, there is also an exhaustive review of theevidence for glaciation in the Mesozoic and Ceno-

.zoic, by Markwick and Rowley 1998 , which con-cludes that there is no convincing evidence forwidespread continental glaciation at any time during

.the Mesozoic. Erikson and Pindell 1998 , in thisbook, demonstrated that the relative sea-level changesthat they can document in northern South Americaoccur over time intervals longer than would be ex-

.pected for glacioeustatic control )1 Ma , but that afew may be eustatic in origin. They compiled a total

of six possible supraregional, potentially global eu-.static events in the Cretaceous, at an average spac-

ing of 13.3 Ma. This number is close to what Miall .1992 estimated would be the case in the Cretaceousrecord, and is a marked contrast to the conclusions of

.the de Graciansky et al. 1998 book.The most recent expression of opinion by adher-

ents of the complexity paradigm can be summed up

in these quotations from Dewey and Pitman 1998,.pp. 1314 :

We cannot conceive of a scenario that wouldallow the synchronous global-eustatic frequencyand amplitude implied by third-order cycles . . . Anespecial problem of using third-order sequencecycles as indicators of global-eustasy is the cor-

.relation problem. Miall 1992 has shown that77% correlation with the Exxon Chart can beachieved between four random number-generatedsequences. Whereas a 0.5-Ma precision is neededfor correlation of the most closely spaced third-order events, a precision of only 2 Ma can beachieved biostratigraphically . . . It is our con-tention that local or regional tectonic control ofsea-level dominates eustatic effects, except forshort-lived glacio-eustatic periods and, therefore,that most sequences and third-order cycles arecontrolled tectonically on a local and regionalscale rather than being global-eustatically con-trolled. We suggest that eustatic sea-level changesare nowhere near as large as have been claimedand have been falsely correlated and constrictedinto a global-eustatic time scale whereby a circu-lar reasoning forces correlation and then uses thatcorrelation as a standard into which all othersequences are squeezed.

Most recent textbooks on stratigraphy reach simi-lar conclusions e.g., see Hallam, 1998, pp. 427428;

.Nichols, 1999, p. 288 . .Kuhn 1962, p. 93 has argued that Awhen

paradigms enter, as they must, into a debate about


paradigm choice, their role is necessarily circular.Each group uses its own paradigm to argue in thatparadigms defense.B He has also stated that twoscientific schools in disagreement will Atalk througheach other when debating the relative merits of their

.respective paradigmsB Kuhn, 1962, p. 108 . Whileaspects of his argument hold true here, the examina-tion of citation practices suggests that the complexityparadigm is more open to external influences in thatit directly addresses the contributions and anomaliesarising from the global-eustasy model. The global-eustasy model, however, adheres to a circular patternof observation and justification, with little attempt toaddress anomalies arising from research outside thecircle of researchers using it. The extent to whichthis practice will continue to elicit support for the

.paradigm is questionable. As Kuhn 1962, p. 93 hasobserved, Awhatever its force, the status of the circu-lar argument is only that of persuasion. It cannot bemade logically or even probabilistically compellingfor those who refuse to step into the circle.B

In Section 8, we address the tenets of the global-eustasy paradigm directly and consider its currentstatus in the scientific community.

8. The global-eustasy paradigma revolution introuble?

.According to Kuhn 1962, p. 93 , Aas in politicalrevolutions, so in paradigm choicethere is no stan-dard higher than the assent of the relevant commu-nity.B Indeed, the most important component of aparadigm is the cognitive authority it is accorded, anauthority based on the judgment of the scientificcommunity that it deserves acceptance over other

.paradigms cf. Gutting, 1984 . We now examine thejudgment of the scientific community as it pertainsto the global-eustasy model itself, with particularreference to the recent book by de Graciansky et al. .1998 discussed in Section 7.

The global-eustasy paradigm is built on the fol-lowing assumptions, or hermeneutic prejudgments:

.1 That a global-eustatic signal exists; .2 That eustatic control generates synchronous

sequence boundaries; .3 That synchroneity can be demonstrated using

current chronostratigraphic tools.

Those scientists skeptical of the global-eustasymodel would argue that none of these three assump-tions has been proven satisfactorily. Indeed, all threehave been the subject of vigorous debate. For exam-

7ple, while the existence of a low-frequency 10 9 .10 -year cyclicity eustatic signal is now generally

acknowledged, there is no agreement on the exis-tence of a high-frequency global signal during peri-

ods lacking widespread continental glaciation e.g.,.during the Cretaceous . But it is not the purpose of

this discussion to repeat the debate about these con-cepts and assumptions. Suffice it to note the follow-ing.

.1 The existence of a suite of eustatic cycles ofAthird-orderB magnitude, and global in extentthebasis for Peter Vails first global cycle chart, hasnever been satisfactorily documented or proven .Miall, 1992, 1997 , a crucial point largely ignoredin the de Graciansky et al. book, in which the Haq et

.al. 1987, 1988 global cycle chart is taken as thestarting point of their work.

.2 It has been pointed out that where relative sealevel is controlled by two or more processes of

similar frequency and amplitude e.g., thermal subsi-dence and second-order eustasy, or foreland basin

.tectonism and Milankovitch cyclicity , a syn- .chronous regional let alone global signal cannot be

generated Parkinson and Summerhayes, 1985; Mi-.all, 1997 .

.3 As noted earlier, conventional chronostrati-graphic methods and results have been subordinatedto the global cycle chart by the Exxon group, andindependent efforts to chronostratigraphically cali-brate the sequence record, such as those by Hallam .many papers, starting with Hallam, 1978 , havebeen largely ignored by the Exxon group.

.4 It has been argued that conventional chronos-tratigraphic techniques are inadequate to test the

reality of high-frequency global synchroneity Miall,.1994 , except in the late Cenozoic, for which the

techniques of cyclostratigraphy calibration of thehigh-frequency sequence record by reference to the

.known frequencies generated by orbital forcing arenow providing a very refined time scale House and

.Gale, 1995 . Graphic correlation methods maychange this, as noted in Section 9.

.4 Several workers have demonstrated the inher-ent diachroneity of sequence boundaries, reflecting


the integrated effects of varying local to regionalsubsidence rates and rates of sediment supply Jordan

.and Flemings, 1991; Catuneanu et al., 1998 .Although the members of the global-eustasy

school refer to Atectonic enhancementB of uncon- .formities Table 2 , at the same time they effectively

rule out consideration of the tectonic and other ef-fects on sequence timing referred to in the previous

.paragraph. As Vail et al. 1991, p. 638 have stated:

structure tends to enhance or subdue eustaticallycaused sequences and systems tract boundaries,but does not affect the age of the boundaries whendated at the minimum hiatus at their conformableposition.

Even where studies appear to have been specifi-cally undertaken to explore the relationship betweensedimentation and tectonics, adherents of the global-eustasy school still regard correlation with the globalcycle chart as significant. Thus, Deramond et al. .1993 , a study not part of the SEPM volume underexamination in this section, explored the relationshipbetween sedimentation and tectonism in tectonicallyactive Pyrenean basins, and succeeded in developingtectonic models for sequence development. How-ever, they then offered correlations of sequenceboundaries with the global cycle chart and went onto discuss eustatic control. In referring to Atectoni-cally enhancedB unconformities, they stated Athe ap-parent correlation between the two groups of inde-pendent phenomena is an artifact of the method,which calibrates the tectonic evolution by compari-son with eustatic fluctuations.B This form ofAcalibrationB is, of course, circular reasoning.

.Hardenbol et al. 1998 summarized the evidenceused in the construction of a set of new chronostrati-graphic cycle charts for the Mesozoic and Cenozoicthat appeared in SEPM Special Publication 60. Theynoted that there are 221 sequence boundaries in thisnew synthesis, in contrast to the 119 that appeared in

.the chart published by Haq et al. 1987, 1988 . Thisis a boundary, on average, every 1.12 Ma, in contrastto the average spacing of 2.08 Ma in the earlierchart. A spacing of 2.08 Ma in the earlier chart wasalready below chronostratigraphic resolution for mostof the Mesozoic and Cenozoic. A potential error of"0.5 Ma is reasonable for the Cenozoic and parts ofthe Cretaceous, but potential errors of several mil-

lions of years must be factored into assigned ages for .older parts of the Mesozoic record Miall, 1994 .

Given a potential error of "0.5 Ma many eventsthose spaced at 1 Ma or lesswould overlap eachother and would therefore be indistinguishable.

Even that earlier chart contains more putativesea-level events than any others published for theMesozoic and Cenozoic, leading to the question:why does the Exxon global cycle chart contain somany more events than other sea-level curves? The

question is even more apposite now. Miall 1997, p..320 suggested that the paradigm of stratigraphic

eustasy followed by the Exxon group allows them tointerpret virtually all sea-level events as eustatic in

.origin. Hallam 1992, p. 92 said, of the Vail and .Todd 1981 article on the North Sea: AThe very title

of the Vail and Todd paper implies that, at that timeat least, seismic sequence analysis of only one regionwas believed to be sufficient to obtain a globalpicture.B Therefore, every new sea-level event that isdiscovered, can be added to the global chart.

This points to a different question: what does theExxon chart actually represent? It undoubtedly con-stitutes a synthesis of real stratigraphic data fromaround the world. It has been suggest that most ofthe sequences are the product of regional events thatoriginated as a result of plate-margin and intraplatetectonism, including basin loading and relaxation andin-plane stresses. These may have been well docu-mented in one or more basins in areas of similarkinematic history, but their promotion to globalevents should be questioned. Some events may beduplicates of other events that have been miscorre-

lated because of chronostratigraphic error Miall,.1992, p. 789 .

Exactly the same assumptions and methodologyappear to have been used in the construction of the

.new charts in de Graciansky et al. 1998 . Sea-levelAeventsB and the corresponding sequence boundariesappear to have been added to the synthesis chartseven when the evidence from separate regional stud-ies in this book indicates that many do not appear ineach of the studied basins. Yet, they are all labeledAglobal.B This has been done despite the clearlystated problems of correlating between Boreal andTethyan faunas within Europe, and the use of North

.American Western Interior chronostratigraphicstandards where European data are inadequate, which


potentially introduces still further sources of errorand imprecision. Finally, because these sequenceshave been documented only in Europe, they cannotbe defined in a blanket manner as having global .eustatic significance.

A logical extension of this approach would be forthe eustasy school to continue to carry out similarresearch in other parts of the world and to continueto add sequence boundaries to their AglobalB synthe-sis. We would predict a resulting chart containingmany hundreds of events, and an average spacing inthe 104105-year range. This may well be what thisteam has in mind, because, as noted below, theycontinue to hold to the mechanism of glacioeustasyas the main cause of most of their sea-level events,and adding still more events would quickly bring theaverage event spacing down to within the Mi-lankovitch band, the normal glacioeustatic fre-quency.

A critical question that needs to be asked in thiscontext, but has not been, concerns scientificmethodology: how can the authors know that a AnewB .as in newly observed sea-level event is in fact atrue AnewB event and not a re

3d1405139879_218213b8c72c0a6107510b50bdde76d3

Documents

globaleustasy model

new paradigm

model of global

theglobaleustasy paradigm

concept of global

effects of eustasy

global changes of sealevel

new ideas