Do Child Abuse Pediatricians Search for a “Pediatric Vulcan Planet”? Comparison of Controversies about the Vulcan-Must-Exist-Theory and the Infant-Must-Have-Been-Shaken-Theory

We argue that there are similarities between the Vulcan-must-exist-theory, derived from the Original Unrestricted Newtonian Gravitational (OUNG) theory, on the one hand, and on the other hand the infant-must-have-been-shaken-theory, derived from the Original Unrestricted Abusive Head Trauma (OUAHT) theory. Although the Vulcan-must–exist-theory was apparently supported by observations over a period of 50 years, after the introduction of Einstein’s general relativity theory in 1915 and its corroboration in 1919, the alleged planet was subsequently neither observed nor needed. In analogy with the Einstein/Vulcan reasoning, we suggest that the introduction of the non-shaken baby theory by Geddes et al. in 2001-2004 indicates that in cases where an infant displays no external signs of trauma, the infant-must-have-been-shaken-theory is no longer needed. Moreover we argue that the two new theories -Einstein’s and Geddes et al.’shave relevant similarities in terms of the effect on the respective original, unrestricted theory. Just as acceptance of Einstein’s general relativity theory led to the abandonment of the Vulcan-must-exist-theory, it is reasonable to claim that the infant-must-have-been-shaken-theory should also be abandoned. We finally argue that


Introduction
The aim of the present paper is to compare and analyze similarities and differences between two apparently very different research areas: astrophysics and pediatrics, with special reference to two theories: the Vulcan-must-exist-theory and the infant-must-have-been-shaken-theory. Our thesis is that if there are relevant philosophy of science similarities between the two theories then they should also be treated equally with respect to promotion or abandonment.
Within astrophysics we focus on the development of the Original Unrestricted Newtonian Gravitational (OUNG) theory and the derived Vulcan-must-exist-theory. The introduction of Einstein's general relativity theory showed that the OUNG theory was no longer unrestricted and that the derived Vulcan-must-exist-theory was actually false and unnecessary. Within pediatrics we focus on the Original Unrestricted Abusive Head Trauma (OUAHT) theory and the derived infant-must-have-been-shaken-theory. In analogy, the introduction of the non-shaken baby theory by Geddes et al. showed that the OUAHT theory is also no longer unrestricted and that the derived the infant-must-have-been-shaken-theory is actually false and unnecessary, at least in cases of alleged Abusive Head Trauma (AHT) without external signs of trauma.
In comparing the two research areas we shall present the interaction between theories and observations, the appearance of anomalies, paradoxical phenomena, auxiliary and ad hoc hypotheses and closure clauses, as well as various associated crises. With respect to associated crises, we focus on three types: fundamental, senility and responsibility crises. We also present the two theories-Einstein's general relativity theory and the non-shaken baby theory by Geddes et al. We illustrate how both theories impregnate observations and interpretations. Moreover, we argue that the two theories have relevant similarities in their effect on the original unrestricted theories. As acceptance of Einstein's general relativity theory meant that the Vulcan-must-exist-theory became superfluous and was eventually abandoned, we argue that it is reasonable to claim that the infant-must-have-been-shaken-theory is superfluous and could also be abandoned in alleged AHT cases without external signs of trauma. We finally argue that while abandoning the Vulcan-must-exist-theory 163 Observation of the curvature of light around the sun during a solar eclipse confirmed Einstein's theory, and the curved spacetime that was found around the Sun explained the "errant" motions of Mercury (Baum & Sheehan, 1997;Levinson, 2015;Kennefick, 2019). Although it was the gravitational pull from Neptune which explained Uranus' odd movements, it was not gravitational pull from the theoretically plausible planet Vulcan that explained the apparently odd movements of Mercury (Baum & Sheehan, 1997;Levinson, 2015).
When Einstein presented his general relativity theory in 1915, he also made a prediction of the magnitude of the deflection of light by the sun-1.75 arc-seconds. In 1919, in conjunction with a solar eclipse, two English astronomers, Arthur Eddington and Frank Dyson, were able to observe the deflection of light and concluded that it was in accordance with Einstein's prediction (Kennefick, 2019).
Although astronomers corroborated Einstein's general relativity theory by observations and prominent (mathematically oriented) physicists approved the theory, there were many antagonists (Kennefick, 2019). The resistance to Einstein's general relativity theory was, in part, due to the difficulty many scientists and philosophers had in questioning the OUNG theory. Initially, Newton had united the terrestrial and celestial physics and the OUNG theory was supposed to apply to all kinds of physical bodies (Kennefick, 2019). Moreover, most physicists thought that light was unaffected by gravitation.
It should also be noted that Eddington actually had problems with the solar eclipse observations. It was cloudy on the small island, Príncipe (off the west coast of Africa) where he made the observations. At the second observational location, Sobral in Brazil, there were problems with calibration of the instruments. However, after careful scrutiny of all the astrographic plates, Eddington and Dyson concluded that the observations corroborated Einstein's theory and their interpretation was endorsed by the Royal Astronomical Society in London (Levinson, 2015, Kennefick, 2019. A computerized reanalysis of the 1919 astrographic plates, published in 1979, showed that Eddington's and Dyson's interpretations were actually correct (Harvey, 1979).
It is important to emphasize that although the OUNG theory was no longer unrestricted, the theory almost perfectly explained the odd movements of Uranus and predicted the position of Neptune. So, despite the fact that the OUNG theory is no longer unrestricted, it is still valid in certain settings where observations are not disturbed by huge gravitational forces.

Astrophysics and Pediatrics-Comparable Scientific Areas?
What is the resemblance between, on the one hand, Newton's original unrestricted gravitational theory making it possible to predict the position and existence of the planet Vulcan, and, on the other hand, the OUAHT theory predicting that an infant must have been violently shaken, based on certain symptoms 166 and signs but without external signs of trauma? Like the OUNG theory, the OUAHT theory presupposes certain initial conditions, clauses and several auxiliary hypotheses (see Box 1)-something we will return to in the discussion of fundamental crises.
Of course we must acknowledge the many and substantial differences between physics and medicine. A common denominator might be that the point of departure in the development of both theories was derived from observations or observational studies. With respect to astrophysics, both Newton and Einstein developed their theories by means of observations, but also applied thought experiments and mathematical reasoning, building upon others physicists' theories. But thought experiments are also based on observations or observational assumptions and observational studies are always of great importance when corroborating or rejecting theories. Einstein's general relativity theory is complicated, much more complicated than the theory to which it is compared in this paper. Currently, however, both within astrophysics and medicine, researchers consider it essential to have observational support for their theories.
Accordingly, it is not remarkable that when developing theories, OUAHT research also takes as its point of departure observational studies and case studies. The latter procedure was applied when the OUAHT theory was developed in the 1970's (Guthkelch, 1971;Caffey, 1972). The OUAHT theories can, however, also be based on microscopic observations of brain tissue from series of fatal cases of alleged AHT infants with and without external signs of trauma . Eventually, medical theories might be corroborated empirically by other scientists who conducted observational studies and reproduced the results. Such studies may then be compiled in meta-analyses or systematic reviews (Maguire, Kemp, Lumb, & Farewell, 2011;Piteau, Ward, Barrowman, & Plint, 2012). On the other hand, theories can be questioned or rejected if the results cannot be reproduced and/or if the evidence is biased (Donohoe, 2003;Elinder, Eriksson, Hallberg, Lynøe, Sundgren, Rosén, ... Erlandsson, 2018;Acres & Morris, 2014).

Our Thesis
Our thesis is that there are some interesting philosophy of science similarities between the past Newton-Einstein controversy and the current controversy over different mechanism theories within the field of AHT research. We suggest that the weakness of the original unrestricted Newtonian gravitational theory was the claim that it represented a solution to all issues and conditions involving gravitational forces. In analogy, we suggest that the weakness of the original unrestricted abusive head trauma theory has been claiming that it is applicable to all alleged AHT cases, including those without external signs of trauma. In order to support these erroneous claims, it was necessary to add the Vulcan-must-exist-theory and the infant-must-have-been-shaken-theory. We argue that these theories are merely false closure clauses (see Box 1) (Johansson & Ceteris Paribus Clauses, Closure Clauses and Falsifiability, 1980 This more recent mechanism theory was presented in the years Geddes, Vowles, Hackshaw, Nickols, Scott, & Whitwell, 2001;Geddes, Tasker, Adams, & Whitwell, 2004). According to this non-shaken baby theory it is not reasonable to claim that alleged AHT cases without external signs of trauma must have been shaken violently , Geddes, Vowles, Hackshaw, Nickols, Scott, & Whitwell, 2001Geddes, Tasker, Adams, & Whitwell, 2004; (Lynøe, Juth, & Eriksson, 2019). We suggest that the infant-must-have-been-shaken-theory can be compared to the Vulcan-must-exist-theory within astrophysics, as applied during the years 1915-1929. Metaphorically we suggest that pediatricians dealing with child abuse currently behave as if they have observed "the pediatric planet Vulcan" and still claim to conduct observations supporting its existence.
But before we further explore relevant similarities and differences between the OUNG theory and the derived Vulcan-must-exist-theory on the one hand and the OUAHT theory with the derived the infant-must-have-been-shaken-theory on the other, we present some key aspects of the development of the latter.

The Development of the Original Unrestricted AHT Theory
In 1962 the American pediatrician Henry Kempe and co-workers presented a study concluding that cases of child abuse were often overlooked by pediatricians, indicating underreporting (Kempe, Silverman, Steele, Droegemueller, & Silver, 1962). Apart from fractures, soft tissue swelling and bruises, the authors also suggested that a certain type of intracranial bleeding -subdural hemorrhage (SDH)-was indicative of child abuse. They also stated that SDH was often missed (Kempe, Silverman, Steele, Droegemueller, & Silver, 1962), indicating that child abuse was underreported. The paper had a huge impact in drawing pediatricians' attention to the issues; thus in 1971 the road was already prepared when the neurosurgeon Norman Guthkelch presented his hypothesis that shaking a baby in a whiplash manner might cause bilateral SDH without external signs of trauma (Guthkelch, 1971). 168 Guthkelch's hypothesis was further elaborated on by the American pediatrician John Caffey who reversed the hypothesis and suggested (abductively) that the finding of SDH and retinal hemorrhages (RH) indicated that the infant had been shaken (Caffey, 1972). According to Caffey, these signs were often observed without external signs of impact to the infants' head. Signs of encephalopathy were later added and the three symptoms and signs-SDH, RH and encephalopathy were referred to as "the triad" (Squier, 2011). The presence of the triad without external signs of trauma has been -and still isregarded as a very strong indication of violent shaking. In 2018 the triad without external signs of trauma was even described as a "smoking gun" (Brook, 2019).
Experimental studies using bio-mechanical models were conducted in order to examine the forces needed to develop the triad, and indicated that high energy forces are a precondition (Duhaime, Gennarelli, Thibault, Bruce, Margulies, & Wiser, 1987;Duhaime, Gennarelli, Sutton, & Schut, 1988;Duhaime et al., 1992;Prange, Coats, Duhaime, & Margulies, 2003). According to these studies, direct impact on the infant's head will exert ~50 times more force than isolated shaking, indicating that shaking alone might be insufficient for development of the triad (Duhaime, Gennarelli, Thibault, Bruce, Margulies, & Wiser, 1987;Duhaime, Gennarelli, Sutton, & Schut, 1988;Duhaime et al., 1992;Prange, Coats, Duhaime, & Margulies, 2003). Nevertheless, it has been claimed that violent shaking can cause and is a precondition for developing: a) SDH, supposedly caused by rupture of bridging veins (Squier, 2011), b) RH, supposedly caused by cleavage between the vitreous body and the retina (Squier, 2011;Levin, 2009), and c) encephalopathy, supposedly caused by brain injury where nerve fibers are torn apart, resulting in immediate symptoms, thus leaving no room for a so-called lucid interval between the suspected violent shaking and the onset of symptoms (Squier, 2011, De Leeuw et al., 2013. Symptoms of encephalopathy include seizures, vomiting, dyspnea, lethargy and unconsciousness. These three events (a-c) associated with findings of the triad are supposed to take place separately and independently, but are all supposedly caused by the same action: violent shaking. Some studies reported that encephalopathy was rarely observed (Vinchon, de Foort-Dhellemmens, Desurmont, & Delestret, 2010), so the presence of solely SDH and RH (i.e., a "dyad") has been considered sufficient to conclude that an infant must have been shaken violently, including cases without external signs of trauma (Vinchon, de Foort-Dhellemmens, Desurmont, & Delestret, 2010).
As suggested in biomechanical studies, violent shaking alone is not sufficient for development of the triad (Duhaime, Gennarelli, Thibault, Bruce, Margulies, & Wiser, 1987;Duhaime, Gennarelli, Sutton, & Schut, 1988;Duhaime et al., 1992;Prange, Coats, Duhaime, & Margulies, 2003). In 2020, a Swedish study showed that isolated shaking -even violent shaking-did not result in triad findings in most of the witnessed or video-recorded abused infants (Thiblin, Andersson, Wester, Wikström, Högberg, & Högberg, 2020). Deceleration-forces to the infant's head are also needed, which may result, for example, from a fall or from the impact of the infant's head on a hard surface, resulting in bruises, ligament injuries and/or fractures. In order to discriminate between accidental and intentional (abuse) injuries, certain criteria were developed-also referred to as AHT-predicting tools (Duhaime et al., 169 1992;Thiblin, Andersson, Wester, Wikström, Högberg, & Högberg, 2020;Reece & Sege, 2000;Cowley, Maguire, Farewell, Quinn-Scoggins, Flynn, & Kemp, 2018). If an infant presents with the triad and the caregiver claims that the infant fell from a height of less than 3 feet, the caregiver is presumed to be lying. Such a fall is considered trivial and not enough to develop the triad-despite the fact that there are cases where the triad has resulted from witnessed falls from a low height (Gardner, 2007). The caregiver who was with the infant when the symptoms were observed is usually identified as the perpetrator, and some studies focus on identifying the potential perpetrator (De Leeuw et al., 2013;Reece & Sege, 2000;Hettler & Greenes, 2003).
These mechanism theories and the associated criteria are important not only to pediatricians dealing with child abuse and child protection teams making the AHT diagnosis, but also to pediatric researchers who use determination of cases by child protection teams to classify cases versus controls in observational studies (Elinder, Eriksson, Hallberg, Lynøe, Sundgren, Rosén, ... Erlandsson, 2018). Such studies are in turn used to corroborate mechanism theories about the triad findings

Corroborating Theories with Observational Studies
When Le Verrier predicted the position of Neptune and later of Vulcan he sought observational support for his predictions. For Neptune, Le Verrier rapidly obtained support from an independent German astronomer (Baum & Sheehan, 1997;Levinson, 2015). In the search for Vulcan he also sought observational support, but this was initially provided by a local amateur astronomer. Le Verrier's prediction of the position of Vulcan was not as exact as that of Neptune. Nevertheless, over a period of 50 years, dark spots on the solar disk were claimed to have been observed many times by a number of observers-both amateur and professional (Baum & Sheehan, 1997;Levinson, 2015).
In OUAHT research, the observational studies involve epidemiological studies including cohort, case-control, cross-sectional as well as diagnostic accuracy studies. Through a significant number of observational studies, the OUAHT theory has obtained empirical support, particularly in meta-analyses (Maguire, Kemp, Lumb, & Farewell, 2011;Piteau, Ward, Barrowman, & Plint, 2012): hence the OUAHT theory appears to be corroborated and robust. However, the individual clinical studies included in the meta-analyses have been criticized. In particular, a systematic literature review disclosed that the large majority of the individual studies of alleged AHT cases without external signs of trauma are at high risk of bias due to circular reasoning (Lynøe, Elinder, Hallberg, Rosén, Sundgren,  However, the systematic literature review (Lynøe, Elinder, Hallberg, Rosén, Sundgren, & Eriksson, 2017) and its criticism of the circular reasoning issue have also been repeatedly criticized, and in a so-called consensus statement the researchers concerned stated what they consider to be the state of art of AHT literature, including AHT cases without external signs of trauma (Choudhary et al., 2018).

The Development of a Scientific Controversy
A scientific controversy is defined as "a persistent antagonistic discussion about a disagreement concerning a substantial scientific issue that is not resolvable by standard means of the discipline involved" (Freudenthal, 2000). As long as everybody embraced the OUNG theory and the derived  . They concluded that "the subdural and retinal bleeding in such cases may well have a physiological cause, rather than being caused by trauma". In other words, violent shaking was no longer necessary to explain triad cases without external signs of trauma-so the infant-must-have-been-shaken-theory was no longer needed in these cases.
It is important to recall that the OUAHT theory might be correct and applicable in the approximately two thirds of all alleged AHT cases where triad findings are accompanied by external signs of trauma . External signs of trauma can implicate a high energy force which can cause disruption of nerve fibers and immediate onset of symptoms (Duhaime, Gennarelli, Thibault, Bruce, Margulies, & Wiser, 1987;Duhaime, Gennarelli, Sutton, & Schut, 1988;Duhaime et al., 1992;Prange, Coats, Duhaime, & Margulies, 2003). Furthermore, high impact trauma might also result in disruption of bridging veins and SDH (Squier, 2011). The primary question is thus whether the OUAHT theory is applicable to the remaining one-third of alleged AHT cases without signs of external trauma .

Comparison of the Vulcan-Must-Exist-Theory and the Infant-Must-Have-Been-Shaken-Theory
In the following we compare the Original Unrestricted Newtonian Gravitational (OUNG) theory and the derived Vulcan-must-exist-theory with the OUAHT theory and the derived infant-must-have-been-shaken-theory. We compare the two theories in the light of more recent competing theories-Einstein's general relativity theory and the non-shaken baby theory by Geddes et al. The two original unrestricted theories and their derived theories are compared with respect to different types of crises, the signs and symptoms of such crises, as well as the use of ceteris paribus-and closing clauses (see Box 1) and different auxiliary hypotheses and ad hoc hypotheses.
Our point of departure is to compare the following two statements: The observation of the odd movements of Mercury prove -according to the OUNG theory-that Vulcan must exist, independent of what astronomers have actually observed.
The observation of the triad (also without external signs of trauma) proves -according to the OUAHT theory-that the infant must have been violently shaken, independent of the caregiver's narration.
In order to scrutinize further our thesis about the analogies between the OUNG theory and the OUAHT theory, let us first focus on the implications of the two more recent alternative theories: Einstein's general relativity theory and the non-shaken baby theory by Geddes et al.

The Implications of Einstein's Theory
Despite the fact that Einstein's theory was corroborated by Eddington and Dyson in 1919, observations of Vulcan continued for at least another ten years. The Vulcan-must-exist-theory was, however, eventually rejected and abandoned after a solar eclipse in 1929 when astronomers found no sign of 172 Vulcan or any other objects inside the mercurial orbit. Most researchers realized that searching for Vulcan was a dead end: scientific interest in Vulcan faded away, as did the scientific controversy (Baum & Sheehan, 1997;Levinson, 2015;Kennefick, 2019). The effect of Einstein's theory was also that Newtonian physics was no longer considered to be unrestricted but still valid and applicable when describing gravitation, such as the interaction between Uranus and Neptune, as well as between any bodies that have mass (Levinson, 2015).
Einstein's general relativity theory not only added something new to astrophysics-it also, eventually, resulted in a new and fruitful research area. Einstein's theory led to a new and productive research program, first during the mid-1950's when observation of quasars and pulsars eventually led to the development of theories about black holes (Levinson, 2015;Kennefick, 2019;Thorne, 1995). scientists were negative or ignorant. Pediatric researchers criticized (Punt, Bonshek, Jaspan, McConachie, Punt, & Ratcliffe, 2004;Richards et al., 2006;Jenny, 2014) and still criticize the theory (Choudhary et al., 2018;Jenny, 2014;Strouse, 2016). This non-shaken baby theory has even been banned in UK courtrooms and classified as a "premature hypothesis" (Supreme Court of Judicature

Implications of the Non-Shaken Baby Theory by Geddes et al.
Court of Appeal (Criminal Devision)., n.d.)! Consequently, the non-shaken baby theory has generally been denied or ignored (Choudhary et al., 2018;Strouse, 2016).
When empirical observations were conducted to test Einstein's general relativity theory and the non-shaken baby theory by Geddes et al., the corroborating studies were also received differently.
But even in 1981, Earman and Glymour questioned whether Eddington was biased when he analyzed the photographic plates (Earman & Glymour, 1981). It should also be noted that when Einstein received the Nobel Prize in 1921 it was demonstratively not for his general relativity theory, but for the law of the photoelectric effect (Hansson, 2017). Currently, there might still exist a small number of denialists of the general relativity theory (Hansson, 2017), but today the large majority of physicists find the theory not only corroborated-it has become a very fruitful theory within modern astrophysics (Thorne, 1995).
Einstein's general relativity theory was eventually accepted, but there was also resistance to the theory, for both scientific and non-scientific reasons: there was reluctance to question Newton's gravitational theory and in 1930's Germany there was anti-semitism. In relation to Newton's gravitational theory, Einstein's general relativity theory represented a scientific revolution. In analogy, the non-shaken baby theory might also represent a scientific revolution, at least for triad cases without external signs of trauma. Both theories were initially denied and the criticism of the Eddington-Dyson observations continued even after a computerized reanalysis corroborated their observations (Kennefick, 2019;Harvey, 1979). It was probably the accumulation of observational support and the fruitfulness of the general relativity theory which eventually led to its acceptance by the large majority of physicists-but this lasted at least one whole generation of physicists. The reception of the non-shaken baby theory was rather hostile, even though independent researchers have supported the theory. We do not yet know the fate of this theory; we may have to wait a further generation before the pediatric clinicians and researchers concerned will accept the non-shaking baby theory and abandon the We now turn to different types of crises and types of auxiliary hypotheses which might occur within a scientific research program under circumstances associated with the Vulcan-must-exist-theory and the infant-must-have-been-shaken-theory.

Signs of Fundamental Crises and Use of Different Auxiliary-and ad hoc Hypotheses
The predicted position and movement of Vulcan was rather difficult to corroborate by observations. This in turn might have indicated that the Vulcan-must-exist-theory was inadequate. Under such circumstances it can be reasonable to develop certain auxiliary hypotheses including ceteris paribus clauses, to specify initial conditions when conducting observations, as well as using closure clauses; it might even be reasonable to develop ad hoc hypotheses (see Box 1). Doing so is not dubious or a sign of a scientific crisis-such procedures might be part of the normal scientific process when testing theories. However, auxiliary hypotheses, initial conditions and closure clauses might be true, truth-like, or false. An accumulation of false assumptions can indicate a fundamental crisis.
In the subsequent text we compare the OUNG theory and the derived Vulcan-must-exist-theory with the OUAHT theory and the infant-must-have-been-shaken-theory, with reference to the use of auxiliary hypotheses, initial conditions and closure clauses. The definitions of the concepts used are presented in Box 1.
One ceteris paribus clause about the OUNG theory was that planets or smaller heavenly bodies might interact with Mercury, thus making the Vulcan-must-exist-theory plausible, in analogy with the Uranus-Neptune interaction.
An example of an auxiliary hypothesis was the examination of black spots observed on the solar disk during the period 1762-1852. These black spots were later interpreted as Vulcan-like and the different sizes and colours of the suspected Vulcan-like phenomena might be understood as an initial condition issue-e.g., a refraction phenomenon (see Box 1).
Another example of an initial condition issue is that it was difficult to observe a heavenly body (e.g., Vulcan) or belts of heavenly bodies close to the sun. Moreover, heavenly bodies can move in an irregular manner and under certain occasions become invisible (Harvey, 1979)-an example of an ad hoc hypothesis (see Box 1).
Certain initial conditions were probably related to technical issues associated with the suboptimal astronomical devices available during the period 1850-1919.
A closure clause (see Box 1) related to the OUNG theory was that light was supposed to be unaffected by gravitation-all other relevant conditions that might be influenced by gravitational forces were excluded, e.g., the notion that even light might be affected by gravity (Table 1).
As it was difficult to question the OUNG theory as a universal theory, supporters of the Vulcan-must-exist-theory had three possible options: (i) continue conducting observational studies and hope that they would eventually be able to corroborate the Vulcan-must-exist-theory (Harvey, 1979;Levinson, 2015), or (ii) develop more speculative auxiliary and ad hoc hypotheses that could explain  , 1980), or (iii) classify the odd Mercury movements as a scientific anomaly in the Newtonian clockwork universe and learn to live with these anomalies without rationalizing them with, e.g., ad hoc hypotheses (Johansson & Lynøe, 2008).
A scientific paradigm might be born, live and die with its anomalies, but generally, signs of a fundamental crisis are, among others, that (i) scientific anomalies become more and more problematic, not only quantitatively but also qualitatively, (ii) several auxiliary hypotheses and essentially defensive ad hoc hypotheses are proposed in order to either make exceptions, help or obscure observational anomalies, and (iii) competing theories have been developed-theories that can provide more plausible explanations or undermine what are currently accepted as the basic theoretical assumptions (Johansson & Lynøe, 2008

Anomalies and Auxiliary Hypotheses
In order to protect the OUAHT theory and the infant-must-have-been-shaken-theory from troublesome anomalies, the advocates have developed auxiliary hypotheses, initial conditions, closure clauses and ad hoc hypotheses (Johansson & Ceteris Paribus Clauses, Closure Clauses and Falsifiability, 1980;Lynøe, Juth, & Eriksson, 2019). Necessary auxiliary hypotheses and closure clauses -which can also become ad hoc-include: (i) An isolated triad cannot develop spontaneously = an ad hoc hypothesis. (ii) If known medical conditions or diseases have been excluded, the only remaining cause is traumatic shaking = an ad hoc closure clause. As biomechanical studies have indicated that shaking alone -even violent shaking-is insufficient to cause the triad, an additional, auxiliary hypothesis has been formed: (iii) The only acceptable explanation in isolated triad cases is that the infant has been violently shaken and if this is questioned, then violent shaking has been combined with slamming the infant's head against a padded surface, which does not result in external signs of trauma = an auxiliary hypothesis.
(iv) No unknown causes bring about the isolated triad = an ad hoc closure clause. (v) The incidence of alleged AHT cases is underreported = an auxiliary hypothesis. (vi) If the caregiver can provide no "acceptable" explanation as to why the infant exhibits the isolated triad, the caregiver is hiding something and accordingly lying = an ad hoc hypothesis. (vii) If the caregiver states that the infant was actually shaken, but as an attempt at resuscitation, such a claim is considered impossible. As an effect cannot precede the cause, the caregiver must be lying = an ad hoc hypothesis (Johansson & Ceteris Paribus Clauses, Closure Clauses and Falsifiability, 1980;Lynøe, Juth, & Eriksson, 2019). (viii) An example of an initial condition issue is the use of a child protection team's determination of alleged AHT cases, despite the obvious and well-known risks of circular reasoning and thus high risk of bias ) (see Box 1).

Auxiliary Hypotheses, Initial Conditions and Closure Clauses
The auxiliary hypotheses, the initial conditions and the closure clauses might be true, truth-like, or false (Table 1). Although several of the hypotheses and clauses presented might appear reasonable, they are actually false, e.g., the claim that all known medical conditions are excluded. Does this mean that there are no new conditions or diseases still to be discovered, or that medical conditions which currently have no known etiology or pathogenesis will remain unexplained and become unexplainable forever?
Finally, claims that a caregiver is lying if he/she cannot explain why the infant has developed the triad signs and symptoms may seem reasonable-but only if the OUAHT theory with all its ad hoc hypotheses and clauses are presupposed to be true, or at least truth-like. They are, however, contradicted by biomechanical studies, as well as by several non-shaken baby theories, which indicate that the triad might develop spontaneously, or represents the sequelae of a delivery-associated SDH (Zahl, Wester, & Gabaeff, 2020). In other words, in triad cases without signs of external trauma, no obvious events have necessarily occurred prior to the infant´s collapse. If the caregiver claims that the infant was actually shaken, but only after the onset of symptoms, this might be correct. Accordingly, these auxiliary hypotheses are also false (Table 1).

Three Issues Might Be True, Truth-like, or False; + Indicates How often Hypotheses, Conditions and Clauses Appear in the Present Context. One Issue Can Appear Twice, e.g., Initial Conditions and Closure Clauses Can also Become ad hoc Hypotheses
Vulcan-must-Infant-must-have-  Vol. 3, No. 2, 2020 In summary, we identified new, competing non-shaken baby theories and the existence of a scientific controversy; we have pointed out several anomalies and paradoxical phenomena, as well as false auxiliary hypotheses, false initial conditions and false closure clauses. The presence of the many false hypotheses, conditions and closure clauses indicates that both theories suffer from a fundamental crisis (Table 2). Moreover, both theories illustrate that because of anomalies and the different types of auxiliary and ad hoc hypotheses, clauses and conditions, it is as difficult to falsify a theory as it is to verify it (Johansson & Ceteris Paribus Clauses, Closure Clauses and Falsifiability, 1980).

exist-theory been-shaken-theory
We identified one additional ad hoc hypothesis associated with circular reasoning. However, because the associated tests will not reveal new knowledge, this hypothesis is associated merely with stagnation.
This is in turn a sign of another type of crisis: the senility crisis.

The Senility Crisis
The Hungarian philosopher Imre Lakatos has described how a scientific research program degenerates if the scientists concerned are unable to rethink the program (Lakatos, 1970). Lakatos contrasts such tendencies to progressive and fertile scientific research programs. The Danish philosopher Stig Andur Pedersen, who further developed Lakatos' ideas and the Kuhnian crisis concept (Andur Pedersen, 1976;Kuhn, 1970), suggested that apart from a fundamental crisis about basic theoretical assumptions, it is also possible to identify a senility crisis and a responsibility crisis (Andur Pedersen, 1976).
According to Andur Pedersen, a senility crisis can be characterized by: (i) stagnation and degeneration of creativity and (ii) focusing on defensive strategies. In accordance with Pedersen's reasoning, we have added: (iii) circular reasoning (this can be considered a defensive strategy, i.e., using non-scientific methods to defend the theory).

Stagnation and Degeneration of Creativity
(I) Within a research program means that new ideas and promising theories are rejected, neglected or ignored and not developed or properly tested Lakatos, 1970). This seems to characterize both the Vulcan-must-exist-theory and the infant-must-have-been-shaken-theory. Instead of reflection and reconsideration the focus might be on routine research -"business as usual"-and the creativity will be limited to developing defensive auxiliary hypotheses and ad hoc hypotheses.
Research on both the Vulcan-must-exist-theory and the infant-must-have-been-shaken-theory ignored and ignore, respectively, new theories that might develop new foci, create new frameworks for interpretations and development of methods, research areas and studies. For almost 70 years (1860-1929)-the Vulcan-must-exist-theory, and during 2004-2020 the infant-must-have-been-shaken-theory and the associated research programs seem to suffer from stagnation, not only lack of creativity. Here is a difference-the Vulcan-must-exist-theory was derived from Newton's gravitational theory, whereas the infant-must-have-been-shaken-theory is derived from an unsubstantiated theory, far from the mathematical power of the Newtonian gravitational theory. It is understandable that abandoning Newton's theory was difficult, but why the Kempe-Guthkelch-Caffey based AHT theory still attracts many advocates is difficult to understand. Nevertheless, proponents of 179  Vol. 3, No. 2, 2020 both theories seem to be interested only in research which can support and defend their preferred theory (Baum, R., & Sheehan W., 1997;Levinson, 2015;Kennefick, 2019;Lynøe, Juth, & Eriksson, 2019).

Defensive and Preserving Strategies
(II) Another sign of a senility crisis is that a research program does not focus on issues which are relevant or which are in conflict with the basic theoretical assumptions. Instead, the research program will focus on defending and preserving strategies. Both the Vulcan-must-exist-theory and its associated research program and the infant-must-have-been-shaken-theory research program have provided several defensive and speculative auxiliary and ad hoc hypotheses. Within the Vulcan-must-exist-theory, the claim that the intra-mercurial bodies might be invisible seems to be the ultimate defensive ad hoc hypothesis. If this ad hoc hypothesis were to be taken seriously we would neither be able to support nor question the existence of Vulcan (Johansson & Ceteris Paribus Clauses, Closure Clauses and Falsifiability, 1980). Moreover, if Vulcan or other mercurial bodies were so small (or had other properties) that they could not be observed, the size of Vulcan would not be sufficient to produce the gravitation pull on Mercury and hence explain its odd movements-unless other ad hoc hypotheses were produced.
Under the infant-must-have-been-shaken-theory we identified several defensive auxiliary and ad hoc hypotheses intended to defend and preserve the OUAHT and the infant-must-have-been-shaken-theory, and protect it from criticism (Parks, Sugeman, Xu, & Coronado, 2012). If we assume that it is correct that whatever the caregiver's version of events in isolated triad cases, the caregiver is assumed to be lying unless he/she admits to having shaken the infant, then it will also become impossible to question the infant-must-have-been-shaken-theory.

Circular Reasoning
(III) If the OUAHT theory is applied by a child protection team to diagnose isolated triad cases as AHT cases, and this diagnostic procedure is subsequently applied to classify allegedly true AHT cases in clinical studies, then the infant-must-have-been-shaken-theory will apparently become corroborated, as will the OUAHT theory. Hence, the theory becomes a self-reinforced, self-fulfilling prophecy based on circular reasoning   Vol. 3, No. 2, 2020 theories based on circular reasoning is like the sculptor who claims that the head he is going to carve out of a marble stone is already pre-shaped in the stone; when he carves out the head-Eureka! -he has proved that the head was indeed pre-shaped in the stone (Eddington, 2012). According to Eddington, this kind of reasoning is common in astrophysics. If circular reasoning is accepted and applied in science in the described manner, there is a high risk that the research programs concerned will not create new knowledge and the research results will become futile.
So far there seem to be several similarities with reference to the fundamental crisis and the senility crisis between, on the one hand, the OUNG theory and the derived Vulcan-must-exist-theory and on the other hand, the OUAHT theory and the derived the infant-must-have-been-shaken-theory (Table 2).

Theory and the Derived Vulcan-Must-Exist-Theory, and the Original Unrestricted AHT Theory and the Derived the Infant-Must-Have-Been-Shaken-Theory regarding Signs of Scientific Crises
+ means that the signs of crisis are present; -means that such signs are lacking. We now turn to the last proposed type of scientific crisis: the responsibility crisis.

Responsibility Crisis
A responsibility crisis might occur when the results of a scientific research program have significant and undesirable societal consequences (Andur Pedersen, 1976 (Kennefick, 2019). But this is the only consequence and it does not result in a responsibility crisis.
This is in contrast to the situation in the field of OUAHT research, including the infant-must-have-been-shaken-theory. In 2012 Guthkelch re-evaluated and clarified his 1971 hypothesis that shaking an infant might result in an SDH, as he began to feel responsible for having facilitated the imprisonment of innocent parents (Guthkelch, 2012). Drs Waney Squier, Julie Mack and Martha Cohen might also have suffered from a responsibility crisis when -following the presentation of the non-shaken baby theory by Geddes et al.-they converted from appearing as expert witnesses for the prosecution, to expert witnesses for the defence (Squier, n.d.; Lynøe, n.d.). Like Guthkelch and later Dr Patrick Barnes, they became aware of the legal and societal ramifications that ensued because of the doubts being raised about the diagnostic accuracy of the isolated triad findings for determining that an infant must have been shaken violently (Barnes, 2017).

Responsibility Crisis and Social Consequences
Those who still embrace the infant-must-have-been-shaken-theory may of course also eventually experience a responsibility crisis, once they begin to doubt the theories. A recently published study showed that child abuse clinicians suffer from so-called secondary traumatic stress associated with burnout (Hymel, 2019). Could this burnout be caused by uncertainty and doubt after diagnosing triad cases without external signs of trauma as AHT and the consequences became obvious, such as 182 unnecessary removal of infants, splitting up of families and sentencing innocent caregivers to prison (Guthkelch, 2012;Squier, n.d.)? If the number of wrongful diagnoses and wrongful convictions in isolated triad cases could be minimized, burnout might decrease (Passmore, Hemming, Chancellor McIntosh, & Hellman, 2020;. Another reason for burnout might be the dilemma that child abuse clinicians have a responsibility to protect infants and it is mandated that the welfare of the infant takes precedent. But this precedence leads to a loss of objectivity. The clinicians do not want to feel responsible for having missed an abuse diagnosis which leads to embracing the diagnosis, rather than questioning it and scrutinizing it.
Because the Vulcan theory had no negative societal or legal consequences, no responsibility crisis occurred among the scientists concerned. A responsibility crisis is expressed solely in the infant-must-have-been-shaken-theory. The responsibility issue is probably also more clearly expressed in pediatrics, at least compared to astrophysics. The majority of pediatricians practice as clinicians rather than as scientists, and hence have limited or no scientific experience and are probably unable to identify scientific fallacies. Some pediatricians who actually have scientific training but also work as clinicians, can also have difficulty in transitioning from the clinical to the scientific setting-and even more so in transitioning from the clinical to the legal setting-and accordingly behave as clinicians in all three settings .
Within astrophysics the large majority of the physicists are, on the contrary, scientists and a corresponding problem consequently does not occur in this field.

Discussion
Obviously there are differences between the two scientific areas-the OUNG theory with the derived One reason why circular reasoning in AHT diagnostics was possible was that pediatric researchers accepted child abuse clinicians' determination of alleged AHT cases without external signs of trauma to determine true positive AHT cases-i.e., as a reference test. Child abuse clinicians are not necessarily scientifically trained, which impedes a critical reflective attitude to their own practice and could make it difficult to acknowledge that current knowledge is fallible (Johansson & Lynøe, 2008). A pediatric researcher is supposed to embrace a scientific attitude, including being critical of one's own assumptions and theories. Accordingly, a clinical researcher would have been expected to abstain from accepting the child abuse clinicians' determination of alleged AHT cases as a reference test. But such a scientific approach seems to decrease or disappear when it comes to child abuse, and pediatric researchers have, contrary to e.g., astrophysicists, found it difficult to retreat to the "scientific ivory tower" and discipline their subjectivity .

Philosophy of Science Similarities
Despite the differences, astrophysics and medical science also have several common denominators with respect to philosophy of science. This becomes apparent when comparing, on the one hand the OUNG theory and the derived Vulcan-must-exist-theory with, on the other hand, the OUAHT mechanism theory and the derived infant-must-have-been-shaken-theory. First of all, both cases illustrate that observations, analyses and interpretations are theory impregnated. Moreover, there are several similarities with respect to anomalies and the different applied auxiliary hypotheses, as well as the signs of a fundamental crisis and a senility crisis.
Because of the lack of legal and societal consequences of the Vulcan-must-exist-theory, it is only in the infant-must-have-been-shaken-theory that we can identify a responsibility crisis. We suggest that the legal and societal consequences explain why the non-shaking baby theory by Geddes et al. is not yet accepted, and that the OUAHT theory has not yet been restricted to alleged AHT cases with external signs of trauma and impact.
The non-shaken baby theory was presented in 2001-2004 and introduced new and different explanations of alleged AHT cases without external signs of trauma and impact. Although Einstein's general relativity theory is much more complicated, the function of the non-shaken baby theory was in principle rather similar. The non-shaken baby theory made it possible to rethink the brain injury (as torn nerve fibers) theory as well as the sources of the SDH and the RH. Moreover, this theory also made the infant-must-have-been-shaken-theory superfluous, at least with respect to triad cases without external signs of trauma. The work of Geddes and coworkers also paved the way for a new and fruitful research program for isolated triad cases. Unfortunately, however, the infant-must-have-been-shaken-theory has not yet been abandoned. The possibility to rethink, revise and develop the research program was never taken seriously, but has on the contrary been largely ignored (Choudhary et al., 2018;Punt, Bonshek, Jaspan, McConachie, Punt, & Ratcliffe, 2004;Richards et al., 2006;Jenny, 2014 Vol. 3, No. 2, 2020 Vulcan planet"-but without the support of a Newtonian gravitational theory. The Kempe-Guthkelch-Caffey theory is based merely on anecdotal evidence and assumptions.

The Reception of the Theories
The reception of the non-shaken baby theory by Geddes et al. may be understood as analogous to a scenario in which Einstein's general relativity theory was ignored and/or denied when it was presented in 1915 and thus never corroborated. We suggest that the lack of interest in rethinking and revising the OUAHT theory may be attributable to concerns over the clinical, legal and societal consequences of acceptance of the non-shaken baby theory. If child abuse pediatricians and other clinicians concerned were to admit that the non-shaken baby theory is actually more plausible than the OUAHT theory, at least in triad cases without external signs of trauma, they would probably feel responsible for not having acted in the best interests of the infant; instead they would feel responsible for unwarranted splitting of families, unwarranted removal of infants and siblings, and for having contributed to the conviction of innocent caregivers. This might already present a problem for at least some neuropathologists and probably also some child abuse pediatricians (Passmore, Hemming, Chancellor McIntosh, & Hellman, 2020;. The general relativity theory eventually brought about an interesting and fruitful research program (Kennefick, 2019;Thorne, 1995). Compared to the research related to the infant-must-have-been-shaken-theory, the non-shaken baby theory might also bring about interesting and fruitful new foci and research programs. However, it may be necessary to await a generational change, following the retirement of those who believed that they were acting in the best interests of the child and its family, but were not .

Concluding Remarks
Although astronomy and medicine are quite different research areas, both use observations to corroborate theories. Both theories discussed here illustrate that observations are theory impregnated.
Although the societal consequences are different, there are several similarities between, on the one hand, the controversy between the original unrestricted Newtonian gravitational theory and the derived Vulcan-must-exist-theory, versus Einstein's general relativity theory and, on the other hand, the controversy between the original unrestricted AHT theory and the derived the infant-must-have-been-shaken-theory, versus the non-shaken baby theory. As both the infant-must-have-been-shaken-theory and the non-shaken theory may have considerable legal and societal consequences, it is solely within this area that scientists might suffer a responsibility crisis.
When comparing research about the Vulcan-must-exist-theory and the infant-must-have-been-shaken-theory, both fundamental and senility crises seem to be almost equal, in terms of signs of crisis such as anomalies, paradoxical phenomena, false auxiliary hypotheses, and debatable ad hoc hypotheses.
In analogy with the introduction of Einstein's general relativity theory, which made the Vulcan-must-exist-theory superfluous, and the fact that the latter theory was subsequently abandoned, 185 www.scholink.org/ojs/index.php/jrph Journal of Research in Philosophy and History Vol. 3, No. 2, 2020 we suggest that the introduction of the non-shaken baby theory will make the infant-must-have-been-shaken-theory superfluous and should therefore be abandoned, regarding the alleged AHT cases without external signs of trauma. But because of ethical concerns and a motivation to protect infants -even when no abuse has taken place-we estimate that it will take at least one or two more generations of pediatricians before the infant-must-have-been-shaken-theory will finally be abandoned. Until then, we will likely witness pediatric researchers, assisted by child abuse pediatricians, continuing the futile search for a "pediatric planet Vulcan"-but instead of standing on Newton's shoulders, standing on Kempe's-Guthkelch's-Caffey's shoulders.