Obsessions, ruminations, worries, and other intrusive thoughts
This post should be read alongside the post on compulsive behaviour and rituals. It is aimed at helping clinicians to differentiate between different types of 'intrusive' thought and to assist the diagnostic process. It covers the types of intrusive thoughts that can occur, and tries to summarise the key features of each. It is written from a diagnostic perspective, and recognises that whilst there are often similarities and overlaps, the different forms of thought usually have diagnostic (and therefore treatment) implications.
Table of Contents
ToggleIntroduction: Normal vs abnormal thoughts
Background
One of the key questions that is still not completely resolved is whether the kinds of obsessions seen in OCD are simply extreme forms of normal thinking, or if they represent an abnormal type of belief. To some extent, this is similar to the question of whether delusions in schizophrenia (e.g. a persistent conviction that other people can read your mind) are primarily a type of 'normal' belief that is has become excessive.
These debates invariably incorporate two distinct questions:
- Can we reliably identify 'normal' thoughts and 'abnormal' thoughts?
- Can we explain 'abnormal' thoughts by what we understand about 'normal' thinking?
To try and answer these questions, we need to look first at how these arguments have developed. However, the history of descriptive psychopathology has been based around a recognition that the 'form' of a thought is often more important than the 'content' of the thought, and that abnormal thoughts are easier to differentiate by their quality, rather than their quantity.
A history of the 'continuity hypothesis'
The continuity hypothesis essentially argues that the thoughts seen in OCD are essentially versions of the types of thinking that occurs in most people. Proponents would argue that in OCD these thoughts are simply excessive. The counter argument is that obsessions are a different type of thought, with differing patterns of reasoning, logic, and reliance on different levels of evidence.
Many supporters of the continuity hypothesis have a background in cognitive therapy, where it is obviously important that such thoughts are considered to be within the scope of such therapy. Also relevant is the unproven implication that if obsessions are on the same spectrum as normal thoughts, then you might be able to intervene early and prevent intrusive thoughts from developing into obsessions (and OCD). We don't know if this is the case and we lack the types of long-term studies that would be needed to explore this.
Rachman & De Silva (1978)
In a seminal study from 1978, Rachman and De Silva compared 'normal' and 'abnormal' obsessions. They interviewed eight people with OCD and "up to" 124 non-obsessional subjects. As was the case for lots of psychology research, the 'control' group was a convenience sample of students and local staff. It is important to note that this group had not been screened for mental disorders.
In the study, "...obsessions were defined as repetitive, unwanted, intrusive thoughts of internal origin." The participants were given a questionnaire which asked about "...the presence or otherwise of intrusive, unacceptable thoughts and impulses, their frequency, and about whether or not these could be easily dismissed."
They found that about 80% of 'normal' subjects answered positively to the questions about whether they experienced obsessions. On the basis of this, the authors concluded that:
- "Obsessions, in the form of thoughts and/or impulses, are a common experience. A large majority of people report experiencing obsessions..."
- "The form. and to some extent the content as well, of obsessions reported by non-psychiatric respondents and by obsessional patients are similar."
- "So-called ‘normal’ obsessions are also similar to ‘abnormal’ obsessions in their expressed relation to mood and in their meaningfulness to the respondent."
The authors did note, however, that normal and abnormal obsessions differed in terms of: ability to dismiss thoughts; intrusiveness; intensity; levels of discomfort; frequency; and degree of resistance. Given that these may be key discriminatory factors, it isn't possible to conclude that OCD obsessions are the same as non-OCD 'obsessions'.
Some other problems with this study include:
- The 'normal' group was not screened for psychiatric disorders that might be associated with intrusive and/ or recurrent thoughts. For example, Obsessive-Compulsive Personality Disorder (prevalence of 2%-4%) or Generalised Anxiety Disorder (prevalence of 2%).
- The sample size of the 'obsessional' group was small (N=8) and robust diagnostic assessments were not conducted. We have no idea whether this group was representative of most patients with OCD.
- Some people reported intrusive thoughts (and were counted as positives) but they did not recognise these thoughts as "unacceptable". The under-developed definition of 'obsession' that was used in the study is likely to result in lots of 'false positives'.
Salkovskis & Harrison (1984)
A few years later, Salkovskis and Harrison attempted to replicate the findings of Rachman and De Silva. The sampling was similar to the original study (i.e. it was opportunistic) and comprised of undergraduate polytechnic students (N=112) and student nurses (N=75). No clinical group (i.e. people with OCD) was used.
The authors found that almost 90% of participants reported experiencing 'obsessions' on the basis of the same questionnaire. Similar results were found with regards to levels of distress and how easy the thoughts were to dismiss.
When we look at the frequency that people were experiencing 'obsessions', however, we can see some problems. Only 6% of the sample had obsessions that were experienced more than ten times per day. This means that 94% of the sample had obsessions that had a frequency of ten times per week or less. Therefore, for the majority of participants, whilst they were having some form of intrusive thoughts, their frequency was low and this level of frequency wouldn't meet typical criteria for OCD (at least one hour per day).
This is in contrast to patients with OCD where obsessions are highly frequent, and sometimes constant. If we define obsessions has being very frequent and highly intrusive, the thoughts that these participants were experiencing were not obsessions. We might conclude, therefore, that whilst intrusive thoughts are common, clear obsessions are not as frequent as claimed.
Purdon & Clark (1993, 1994)
In two linked studies, Purdon and Clark reported on the development of a new tool to measure intrusive and other types of negative thought: the Obsessional Intrusions Inventory (OII). The tool was evaluated in 29 undergraduate students. This means that their findings were based on a small group of people who didn't have OCD.
The OII included a range of intrusive thoughts, including: insulting people; running your car off the road; jumping from high places; leaving your home unlocked; and causing a fire/ accident by leaving something on. Therefore, the subjects of included thoughts included areas that may or may not be clearly identified as typical obsessions seen in OCD.
The authors reported that some thoughts were common, typically those related to some degree of social embarrassment and not being careful enough. Being violent towards strangers or family members was much less frequent (5%-10%). The second part of the study explored relationships between the thoughts and the individual's experience of the thoughts.
They concluded:
"Based on the present findings, it is evident that the vast majority of nonclinical individuals report having thought intrusions of sex, violence, dirt and contamination that are also characteristic of clinical obsessions. This thought content is very similar to the type of thinking one encounters in obsessive compulsive disorder."
"...we found that the extent to which individuals believed that their most upsetting intrusive thought could lead to unacceptable behaviour in real life was closely related to both the frequency and persistence of the intrusion as well as its perceived uncontrollability. The stronger one believes that he/she could act on the intrusion, the more persistent and less controllable the intrusive thought."
"...it appears that the characteristic control response of highly obsessional individuals to their unwanted, distressing intrusive thoughts is not that unusual or unique, but rather these individuals experience less success in utilizing these control strategies than low obsessional individuals."
All we can probably say is that most people have some kinds of thoughts that are experienced as intrusive, but that we can't determine whether these were actually obsessions. They may have looked (superficially) like obsessions, but without detailed clinical assessments it is impossible to say. Thoughts associated with social embarrassment or needing to do things properly (e.g. locks and switches) are likely to be common, but the presence of such thoughts does not automatically mean that they are pathological.
Rassin and Muris (2007)
To explore the boundaries between 'normal' and 'abnormal' obsessions, Rassin and Morris presented the typical obsessions identified by Rachman et al to six psychologists and eleven psychotherapists, and also to 90 psychology undergraduates.
They found that:
"Both therapists and students were able to distinguish clinical and non-clinical obsession beyond chance level. It is concluded that some clinical obsessions can be identified as being evidently abnormal"
They also reported that:
"...therapists were not better at discriminating abnormal and normal obsessions than were undergraduate students. This indicates that it is not clinical experience that enables judges to distinguish the two kinds of obsession. It also adds to the claim that, in absence of standardised diagnostic instruments, clinical psychologists hardly outperform the less experienced undergraduate students...If it is not clinical experience, the question remains which criteria enabled the participants to differentiate between abnormal and normal obsessions."
This study did not involve actual patients and simply asked various groups to try and identify 'normal' and 'abnormal' thoughts. Since the various types of thought were not necessarily obtained from clinical or non-clinical populations, it remains speculative as to whether some thoughts were truly 'abnormal'.
'Intrusive thoughts on six continents' (Radomsky, Moulding, and Clark, 2014)
In a series of three papers, Radomsky et al explored the nature and experience of obsessions and other intrusive thoughts in non-clinical populations. They used the International Intrusive Thoughts Interview Schedule (IITIS) to assess prevalence of intrusions in 777 university students at 15 sites in 13 countries across 6 continents. The sample size was large, but it did not include clinical populations. That is, no-one in the study had OCD.
Their starting point was that intrusive thoughts are common, that they probably exist on a continuum with obsessions, and that it's not necessarily the intrusions themselves that are the problem. Rather, it's the misinterpretation and how people respond to them that is problematic. This fits with cognitive 'appraisal' models of abnormal thinking and the development of psychiatric symptoms. One could argue that the study was an attempt to confirm (rather than disconfirm) an established model that most of the researchers undoubtedly had strong allegiances to.
Across these three studies (each describing different aspects of their findings), it was reported that:
- 94% of participants reported at least one intrusion during the previous three months.
- The most common category of intrusions was doubting, whilst some of the more typical OCD obsessions (such as sexual or blasphemous thoughts) were much less common.
- "Repugnant vs. non-repugnant thought-content differed only with respect to thought frequency, but thought-content did not moderate the relationship between intrusive thoughts and control strategies and appraisals."
- "...appraisals and control strategies generally partially mediated the relationship between general OCD-related beliefs and the occurrence of, and distress associated with, intrusive thoughts."
- Normal and abnormal obsessions were continuous.
- Some types of obsession (dirt/ contamination, doubt, and 'miscellaneous') were more specific to OCD.
Unfortunately, whilst these studies confirm that intrusive thoughts (in some form) are common in non-clinical populations, they don't offer clear mechanisms by which we can reliably differentiate between non-clinical 'intrusions' and clinical 'obsessions'. No-one in the study had OCD. What they do suggest, however, is that the frequency of non-clinical intrusions is much less than that seen in patients with OCD. Whilst non-clinical populations may experience one or two intrusive thoughts per week, patients receiving treatment for OCD will typically be experiencing hundreds of intrusive thoughts per day.
Are normal and abnormal intrusions on a continuum?
The answer is that it's probably hard to say. Whilst many studies would support the commonality of many types of intrusive thoughts, most studies have been conducted with a particular worldview in mind, and have rarely involved clinical populations. This means that whilst it's possible to say that most people will experience some form of intrusive thought, we can't say at what point those thoughts are more likely to be classed as obsessions, and whether that depends simply on the frequency or intensity of those thoughts or if there is something inherently different about obsessions. This is a 'quality' (type or form of thought) vs 'quantity' (frequency or intensity) type of question.
Are obsessions different from worries?
It could be suggested that if obsessions are very similar to worries, then the overlap/ co-occurrence between OCD and Generalised Anxiety Disorder (GAD) should be high. Although OCD and GAD can occur alongside each other, the co-occurrence is not that high; probably about 18% (e.g. Welkowitz et al, 2000). We will see later some of the differences between thoughts in OCD vs GAD.
Proposed differences between obsessions and worries
Turner et al (1992) proposed several key differences between obsessions and worries:
- Worries are more ego-syntonic (consistent with the person's underlying belief system) than obsessions. A typical worry might be that you are concerned about a loved-one's health (because you are a caring person), whilst an obsession might involve intrusive thoughts about harming a loved-one (even though you are a caring person).
- The content of a worry is not experienced as being as unacceptable as the content of obsessions. Similar to above, worries about health, finances, or the future seem much more reasonable to people than most obsessions.
- People with worries are often much more aware of internal and external triggers. Worries are more often triggered by real-world (and relevant) events.
- Worries are typically experienced as the person's internal 'voice', whilst obsessions are more likely to be thoughts, images, or impulses.
- Worry is less resisted, and is often less intrusive. However, some people with severe GAD can experience highly intrusive worries.
Similarities between obsessions and worries (Turner, 1992)
Similarities include:
- Obsessions and worry can occur in both clinical and non-clinical populations.
- The content can be similar in clinical and non-clinical population. That is, the things people worry about are not robust distinguishing features to differentiate clinical and non-clinical groups.
- Negative emotions (anxiety, guilt, self-recrimination) can occur alongside both obsessions and worry.
- The main thing that separates clinical from non-clinical groups is frequency and intensity.
Is there experimental evidence to answer the question?
Langlois et al (2000) used the Cognitive Intrusion Questionnaire to examine intrusive thoughts and worry in 254 students. The population was, therefore, not a clinical one. Participants were not screened for psychiatric illness and they took part with the possibility of winning a $50 prize, so this was undoubtedly another 'opportunistic' sample rather than one that might be able to answer the key questions. As a result, some assumptions have to be made, for example:
- That the population of students doesn't have significant psychiatric morbidity. We have seen already that many people in such samples will have conditions that involve excessive worry and recurrent thoughts. That might mean that we can't always assume that such populations are 'normal' (i.e. free of psychiatric symptoms).
- That what is identified as 'obsessions' are actually obsessions. In reality, the authors were assessing intrusive thoughts, which aren't automatically obsessions.
- The Cognitive Intrusion Questionnaire (CIQ) was used to identify/ assess intrusions and the Worry Domains Questionnaire (WDQ) was used to assess worries. We therefore have to assume that everything that the CIQ measures was an intrusion/ obsession and everything on the WDQ was a worry. Without interviewing participants, we can never be sure.
The authors reported that 'obsession-like intrusions' differed from worries on several factors, including: frequency; duration; egodystonia; emotional response; and the amount of verbal and image content. They suggested that whether the thoughts/ worries were egodystonic or egosyntonic was an important discriminatory feature between them.
Definitions of obsessions and worry
In the definitions below, I have intentionally created two categories. One is the type of thought that occurs in OCD (true obsessions) and the other includes other types of intrusive thought, and would include worry, doubts, and ruminations. There will be overlaps between the different types of non-OCD thought, and the choice of term is usually governed by what school of cognitive therapy an author comes from and, to some extent, the time in which the author was writing. There are no robust or reliable definitions of 'worry', 'doubt', or 'rumination' that is universally accepted.
Obsessions
The definition of an obsession below is based around the diagnostic criteria for OCD that have existed for decades. It is important to realise that whilst the majority of obsessions are thoughts, they can also be images or impulses.
The characteristics of an obsession are as follows:
- The thought/ doubt/ image/ impulse is unpleasantly intrusive. This means that it intrudes upon normal thinking and is present in situations where the thought is not relevant.
- It is repetitive. It re-occurs, despite the individual trying to distract themselves or to reassure themselves that the fear is not true.
- It usually generates anxiety and/ or distress. This can sometimes diminish over time, but there is almost always a clear description or a history of the thought causing anxiety.
- It is recognised as coming from within the person's own mind. This differentiates it from other types of psychotic symptoms where thoughts are experienced as 'alien' or originating from outside of the person's own mind.
- The thoughts are resisted. The person will typically try to distract themselves or push the thoughts away.
- They are ego-dystonic. This means that they are not felt as being consistent with the person's longstanding beliefs or personality. For example, someone with strong religious beliefs may develop blasphemous thoughts, or someone with a strong sense of being caring might have thoughts of harming those close to them.
- Obsessions are recognised by the individual as being excessive and/ or unreasonable. It can sometimes be challenging when someone believes the thought to be a reasonable one (e.g. is the door locked?) but they can't dispel the doubts. So they might see the thought as reasonable, but still excessive.
Worry, ruminations, and non-obsessional intrusive thoughts
Worry shares a lot of similarities with obsessions (see above). When trying to provide a definition, we are not making a clear differentiation between 'worry' and 'rumination'. In the literature, the distinction typically depends on who is writing at the time and no clear distinction has been proposed between worry and rumination. Often, they are used interchangeably.
With regards to intrusive thoughts, it's sufficient to say that the category of 'intrusive thoughts' obsessions and worries. However, this doesn't mean that obsessions are the same as worries.
To some extent, the separation of obsession and worry is made on the basis of the 'form' of the thought, rather than the 'content'. This means that other factors (such as frequency and distress) may be more important than simply the content of the worry.
On this basis, worry is usually characterised by the following:
- A worry is a persistent or recurrent thought that typically arises in response to anxiety-provoking events or situations, and is usually associated with other symptoms such as distress or fear about the future. Worries are often increased by depressive symptoms and other types of anxiety.
- More often than not, a worry relates to 'real-world' and rational situations that are relevant to the person's life. For example, this someone might worry about finances, health, the future, making mistakes, not being liked, relationships, being judged, etc. Thoughts about irrational situations (such as stabbing a friend) are less likely to be a worry (although they may be described as such).
- Worries are usually 'egosyntonic'. That means that the worry itself makes sense to the person and the worry is consistent with the person's pre-existing beliefs and personality. Some people might say that they have always been a 'worrier', and they can understand why they might be worrying about something. It 'makes sense' to the person why they might have a particular worry and might say that thinking (or overthinking) about something is a legitimate way to resolve the problem.
- The worry is usually context-specific. This means that people worry more when the thing they are worrying about is more salient (or relevant). So, having lots of bills, losing a job, or a recent health scare might increase worries about these kinds of subjects.
Types of symptom that are less likely to be due to OCD
Ruminations in response to unwanted behaviour
Some people might have a tendency to engage in sexual practices that are commonly seen as problematic and the person may feel bad afterwards, or feel guilty about the behaviour. For example, sex with prostitutes or seeking out particular types of pornography (e.g. under-age). Such ruminations may even be associated with reassurance-seeking. However, there are some key differences:
- These non-OCD ruminations are consequent upon the unwanted behaviour, rather than the behaviour following an unwanted thought. Since the causal relationship is different, despite their repetitive and/ or intrusive nature they are less likely to be obsessions in the traditional sense.
- Importantly, people with OCD are avoidant of situations that they find distressing. Although they may try to 'test out' their underlying beliefs (e.g. am I gay?, am I a paedophile?), they do not willingly seek out stimuli that are pleasurable or desirable. The fear is present throughout any kind of 'testing themselves' behaviour (often along with feelings of shame and/ or disgust) and does not arise only after the behaviour has been completed. People who engage in undesirable (but pleasurable) activities don't show the same levels of avoidance.
Worry in Generalised Anxiety Disorder (GAD)
The types of worry that arise in GAD have been explored above, along with key differences in how the symptoms are experienced.
Worry in obsessive-compulsive (anankastic) personality disorder (OCPD)
People with OCPD will often have worries relating to perfectionism, not getting things correct, or that their inability to do something properly will result in adverse consequences. Due to the overlaps between OCD and OCPD, it can sometimes be difficult to separate such symptoms. There is more discussion of this in the post on compulsive behaviour and rituals, with some specific examples of how OCD and OCPD might differ.
Thoughts in Autism Spectrum Disorder (ASD)
It is quite common for people to have both OCD and ASD. Additionally, they may have thoughts that are a mixture of OCD and ASD, and repetitive behaviours may represent both OCD and ASD and be present at different times or the same time.
In ASD-related thoughts there are often some key differences:
- The thoughts are commonly very 'black and white' and similar to classical patterns of inflexible thinking. There may be some thoughts that have arisen from single experiences, for example: "I felt terrible last time that happened, so I'd better make sure it doesn't happen again."
- The preoccupations are more likely to be about negative experiences associated future events rather than the anxiety associated with current events. The feared future events might be less about explicit outcomes seen in OCD (e.g. "I might make my family unwell by spreading germs") and more along the lines of, "If I don't do this properly, I'll feel bad or something bad might happen."
- Preoccupations may have a strong somatic (i.e. physical quality). Worries about going to the toilet might be more about the process of going to the toilet or how it feels after going to the toilet, and less about clear thoughts about spreading contamination.
- Intolerance to distress is often a mediating factor in how behaviours become linked to thoughts. So, behaviours are performed as 'safety behaviours' and to reduce distress (commonly described as 'self-stimming') rather than being explicitly linked to specific feared outcomes that could exist in the external world.
Comparison between OCD and GAD
The table below summaries some of the above information. No single characteristic is determinant of one type of intrusive thought versus another, but looking at the overall pattern of thought(s) should help in differentiating them.
Key: ++ Strongly characteristic; + Moderately characteristic; - Not strongly characteristic.
Feature | OCD-related obsessions | GAD-related worry |
The thoughts are ego-dystonic | ++ | - |
Duration is more likely to be described as frequent/ constant rather than episodic/ intermittent | ++ | - |
The thought generates anxiety | ++ | - |
The thought arises out of anxiety, or anxiety feels like the 'driving force' behind the experience | - | ++ |
The sense of control over the thought is moderate to high | + | ++ |
The thought content is bizarre, impossible, or irrational | ++ | + |
The thought is resisted, rather than being seen as a suitable way to resolve a situation | ++ | - |
The thought is more likely to be specific ("What if I get HIV and spread it to my wife?") rather than general ("Everything is dirty and unsafe") | ++ | - / + |
The thought is seen as a possible way to 'solve' a problem. For example, "I need to think about this in order to fix a problem" | - | + |
The content of the thought (rather than the situation in which they occur) relates to interpersonal issues (such as relationships, health, work, friendships, etc.) | + | ++ |
The thought involves a possibility becoming reality, usually via far-fetched mechanisms | ++ | - |
The thought is rooted in reality, and the mechanisms are very tangible and real | - | ++ |
The thought occurs 'out of context'. It may be triggered by external factors, but it is not related to current or actual realities. | ++ | + |
References
Rachman, S. & de Silva, P. (1978) Abnormal and normal obsessions. Behaviour Research and Therapy, 16, 233-248. http://doi.org/10.1016/0005-7967(78)90022-0
Volkert, J., Gablonski, T.-C. & Rabung, S. (2018) Prevalence of personality disorders in the general adult population in Western countries: systematic review and meta-analysis. British Journal of Psychiatry, 213, 709-715. http://doi.org/10.1192/bjp.2018.202
Grant, B. F., Hasin, D. S., Stinson, F. S., et al (2005) Prevalence, correlates, co-morbidity, and comparative disability of DSM-IV generalized anxiety disorder in the USA: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Psychological Medicine, 35, 1747-1759. https://doi.org/10.1017/S0033291705006069
Salkovskis, P. M. & Harrison, J. (1984) Abnormal and normal obsessions - A replication. Behaviour Research and Therapy, 22, 549-552. http://doi.org/10.1016/0005-7967(84)90057-3
Purdon, C. & Clark, D. A. (1993) Obsessive intrusive thoughts in nonclinical subjects. Part I. Content and relation with depressive, anxious and obsessional symptoms. Behaviour Research and Therapy, 31, 713-720. http://doi.org/10.1016/0005-7967(93)90001-B
---- (1994) Obsessive intrusive thoughts in nonclinical subjects. Part II. Cognitive appraisal, emotional response and thought control strategies. Behaviour Research and Therapy, 32, 403-410. http://doi.org/10.1016/0005-7967(94)90003-5
Rassin, E. & Muris, P. (2007) Abnormal and normal obsessions: A reconsideration. Behaviour Research and Therapy, 45, 1065-1070. http://doi.org/10.1016/j.brat.2006.05.005
Radomsky, A. S., Alcolado, G. M., Abramowitz, J. S., et al (2014) Part 1. You can run but you can't hide: Intrusive thoughts on six continents. Journal of Obsessive-Compulsive and Related Disorders, 3, 269-279. http://dx.doi.org/10.1016/j.jocrd.2013.09.002
Moulding, R., Coles, M. E., Abramowitz, J. S., et al (2014) Part 2. They scare because we care: The relationship between obsessive intrusive thoughts and appraisals and control strategies across 15 cities. Journal of Obsessive-Compulsive and Related Disorders, 3, 280-291. http://dx.doi.org/10.1016/j.jocrd.2014.02.006
Clark, D. A., Abramowitz, J., Alcolado, G. M., et al (2014) Part 3. A question of perspective: The association between intrusive thoughts and obsessionality in 11 countries. Journal of Obsessive-Compulsive and Related Disorders, 3, 292-299. http://dx.doi.org/10.1016/j.jocrd.2013.12.006
Langlois, F., Freeston, M. H. & Ladouceur, R. (2000) Differences and similarities between obsessive intrusive thoughts and worry in a non-clinical population: study 1. Behaviour Research and Therapy, 38, 157-173. http://doi.org/10.1016/S0005-7967(99)00027-3
Welkowitz, L. A., Struening, E. L., Pittman, J., et al (2000) Obsessive-Compulsive Disorder and Comorbid Anxiety Problems in a National Anxiety Screening Sample. Journal of Anxiety Disorders, 14, 471-482. http://doi.org/10.1016/S0887-6185(00)00034-7
Turner, S. M., Beidel, D. C. & Stanley, M. A. (1992) Are obsessional thoughts and worry different cognitive phenomena? Clinical Psychology Review, 12, 257-270. https://doi.org/10.1016/0272-7358(92)90117-Q
Last Updated on 15 November 2024 by David Christmas