Elsevier

Dendrochronologia

Volume 44, June 2017, Pages 211-214
Dendrochronologia

Short Article
Reconciling the community with a concept—The uniformitarian principle in the dendro-sciences

https://doi.org/10.1016/j.dendro.2017.06.005Get rights and content

Highlights

  • The uniformitarian principle is the basis for extrapolating into the past.

  • The dendro-community uses this principle inconsistently and partially incorrect.

  • We propose a simple, logical and straightforward interpretation.

Abstract

The uniformitarian principle is one of the most important foundations of all dendro- and paleo-sciences. Without it, no inferences about the past can be made. However, the use of this principle in our community is not consistent and partially incorrect, with the main confusion relating to the understanding of the “uniformitarian principle” as somehow implying a stable relationship between climate and tree growth. To solve this, we look briefly at the history of the term, show how we teach this principle in our textbooks, give some examples of incorrect applications of this principle in the recent literature and close with a simple, logical and straightforward interpretation of this principle to the dendro-community. Applying the principle of aggregate tree growth we show that instable climate-growth relationships and the “no-analogue problem” are not a violation of the uniformitarian principle, but rather reflect our incomplete understanding of tree growth processes. Simply stated: The “uniformitarian principle” is an a priori assumption of spatial and temporal invariance of law’s describing nature’s processes. Applied to the dendro-sciences it means that the principle of aggregate tree growth is valid in time and space.

Introduction

Most of us would agree that the science of dendrochronology, and its relatives such as dendroarchaeology, dendrogeomorphology, or dendroecology among others, rest on seven basic principles: The Uniformitarian Principle, the Principle of Limiting Factor, the Principle of Aggregate Tree Growth, the Principle of Ecological Amplitude, the Principle of Site Selection, the Principle of Crossdating and the Principle of Replication (Fritts, 1976, Speer, 2010). We all were taught and are teaching these principles to students and newcomers to the field, in university courses, summer schools, workshops and at conferences. We are basing our daily work on these principles.

The uniformitarian principle is usually stated as the first and is considered the most basic principle. Its importance was highlighted by Hal Fritts back in the first edition of his textbook “Tree Rings and Climate” (1976): “The uniformitarian principle is assumed in all dendrochronological inferences, and, as in all sciences of the past, if this principle does not hold, no conclusions regarding the past can be made.” This principle is often stated simplified as “the present is the key to the past” and is also part of “uniformitarianism” or called “actualism” in other sciences.

However, over the years, the authors have noted that our community uses this term with slightly differing meanings and while on a day-to-day basis this is often neglected − a principle should not leave room for a lot of personal interpretation or risks severe misunderstanding. This communication serves therefore two main purposes: 1) To briefly reiterate the development of the uniformitarian principle from geology to its modern application in dendro-sciences, which is important to understand some disparities, and 2) to propose a simple and straightforward interpretation of this principle to the dendro community.

We do this by looking briefly at the history of the term (“A: The awkward eight syllable word − Uniformitarianism”), show how we teach this principle in our textbooks (“B: The Uniformitarian principle as we teach it”), give some examples of differing interpretations of this principle in the recent literature (“C as Confusion”) and close with a recommendation (“D: The way forward”).

James Hutton (1726–1797) is often credited as the person coining the catchy phrase “The presence is the key to the past”. While this is most likely not true (it probably was Sir Archibald Geikie (1905), when writing about Hutton), he is considered by many as one of the main fathers of modern geology. His dissertation, read in 1785 in front of the Royal Society of Edinburgh, was titled “System of the Earth, its duration and stability” and we have to remember that it was set in a time of geological dispute between Neptunists, Catastrophists and scientists trying to show that the earth was actually older than 6000 years. It is difficult to understand from our viewpoint today, but Hutton seems to have been one of the first paleo-scientists using modern inductive theory and formulating the basis for what was to become known as “uniformitarianism” a little later.

In 1830–1833, Sir Charles Lyell made a mark in geology with a three volume textbook “Principles of Geology”. He constructed a theoretical system to establish geology as a natural science and based his system on three principles (after Camardi, 1999): “1) the Uniformity Principle, which states that past geological events must be explained by the same causes now in operation; 2) the Uniformity of Rate, which states that geological laws operate with the same force as at present; 3) the Steady-state Principle, which states that the earth does not undergo any directional change.” Taken together, these three principles became known as “Uniformitarianism”. Uniformitarians set out to explain the geological record with processes observed in the present day, opposing the “catastrophists” (Scott, 1963) who used “the direct agency of Creative Interference” (Buckland, 1837 in Gould, 1965a), in other words divine intervention, to explain the record. However, Lyell snuck his personal view of “gradualism” into this framework, which implies invariance of rates of change (see principle 2), hampering the science of geology for at least 150 years (Allmon, 1993). Luckily, we don’t have to wade through decades of scientific debate (for those interested see Shea, 1982, debunking the concept), but can jump ahead and take a look at the first scientific paper from Steven Jay Gould, the great American paleontologist, evolutionary biologist and popular science writer.

In 1965 Gould published “Is uniformitarianism necessary?” and answered his rhetorical question with a definite “No”. Gould (1965a; b) clearly drives home the point that the term “uniformitarianism” confounds two different concepts: “A testable proposition asserting constancy of rate of change or material conditions through time – substantive uniformitarianism” (Lyells principles 2 and 3, see above), and “An undemonstrable, though entirely necessary, procedural assumption asserting spatial and temporal invariance of law’s describing natures processes – methodological uniformitarianism” (Lyells principle 1, see above) (Gould, 1965b). He then assessed both concepts and concludes (Gould, 1965b) “the testable proposition is incorrect in any strict formulation” and “the concept embodied in the procedural assumption is essential, but the term ‘uniformitarianism’ is unnecessary, since the invariance assertion is synonymous with the principle of induction”.

While we can follow Goulds logic in dismissing Lyells principles 2 and 3 as rejected hypotheses today, we personally see a value in retaining Lyells principle 1 as a reiterating backbone of our scientific approach, or a specific application of the general inductive concept to paleosciences in general and dendrochronology in particular. Two years before Gould, in 1963, Reyer Hookyaas had discussed Lyells principles and also retained only the first one. When talking about von Hoff, a contemporary of Lyell, he referred to the “Uniformity Principle” (principle 1) of Lyell as “actualism” (Hookyaas, 1963, page 10): “…he hoped to remain true to the principle that causes now in operation are sufficient to explain the ancient volcanic phenomena, that is, he remained true to ‘actualisme’ taken in the literal meaning of admitting none but causes “actually” (at present) in operation…”

This principle, may it be called “Uniformity Principle” or “actualism”, is, in our opinion, closest to the “uniformitarian principle” originally introduced into the dendro-sciences. It is important to note that this principle does not refer to any testable hypotheses and is NOT synonymous with the historical “uniformitarianism” as a concept, but is rather a “procedural assumption‘ (after Gould, 1965b) of using inductive reasoning (“all small rings today in these pines on sand are a result of drought, so a small ring in a pine of the past which grew on sand most likely was the result of drought as well – all other factors being similar’).

After this excursion into history, it is time we take a look at our science. The tools for our trade are written down in handbooks, for reference and studying, and more and more ‘somewhere in the net’. This section briefly summarizes how the uniformitarian principle is defined in textbooks of dendrochronology and on random available webpages, which, let us be honest, students would turn to first, if they had to look up the principle (Table 1). The summary contains only the field of dendro-sciences and not examples from other fields – as our goal is here to show how we, as the dendro-community, go about using the principle. Please note: While most researchers refer to the “uniformitarian principle”, some do use the term “uniformitarianism” (see C as Confusion).

As far as we can see, there are two (possibly related) types of confusion regarding the principle under discussion. The first relates to the difference between “uniformitarianism” (as used historically and still partially today) and the “uniformitarian principle”, and the second to the meaning of the latter.

Most colleagues defined the “uniformitarian principle” and not “uniformitarianism” in their textbooks and teaching material and we would highly recommend to our community to do so in the future. Taking into consideration that our community borrowed the concept from geology, bringing “uniformitarianism” into the mélange can be counterproductive, as it might imply the possibility of constant rates of change (Lyells principle 2), and might give life to a historical concept, which the original science (geology) has outgrown (but see Knight and Harrison, 2014).

The second possible confusion relates to the understanding of the “uniformitarian principle”. Several recent papers (including one of the first author) and textbooks highlight a non-stationary climate-growth response (or changing climate sensitivity), possibly related to a shift in the seasonal signal, as a possible deviation from the uniformitarian principle (e.g. Briffa et al., 1998; Wilson and Elling, 2004; Carrer and Urbinati, 2006; D’Arrigo et al., 2008; Wilmking and Singh, 2008; Leonelli et al., 2009; Büntgen et al., 2010; Speer, 2010; Magda et al., 2011; Trindade et al., 2011; Büntgen et al., 2012; Dorado-Liñán et al., 2012; Zhang et al., 2013; Franceschini et al., 2013; Naulier et al., 2015). Apart from the fact that, as one reviewer mentioned, non-stationary climate-growth relationships can be due to a number of factors such as “declining quality of climate data back in time” or “the way that the tree-ring data have been processed and statistically compared to the climate data”, we infer that many of these authors mentioned above seem to interpret the “uniformitarian principle” as indicating a (historically) often assumed linear (note reference to Lyells principle 2) and stable relationship between a tree ring proxy and climate drivers.

However, if we follow the logic outlined above, and apply a modern understanding of the “uniformitarian principle”, these statements are not true. We claim that there is no deviation from the “uniformitarian principle” in any of the examples cited above. The general laws of nature are still at work and possible non-stationary relationship between climate and tree growth would have also happened in the past given the same circumstances (not only climatic, but also same site conditions, same tree ages, same environmental drivers – see also Visser et al., 2010).

It might be helpful here to ask another of our basic principles to lend a hand, the “Principle of aggregate tree growth” (after Cook, 1987, Speer, 2010, modified sensu King et al., 2013).Rt = f (g1Gt, g2Ct, g3δD1t, g4δD2t, Et)Rt is the ring width in year t, which is influenced by age and size trends (Gt), climate (Ct), endogenous disturbance with a stand (D1t), exogenous disturbance from outside the stand (D2t), the δ associated with D1 and D2 is a binary indicator of the presence or absence of either class of disturbance in the ring widths, g1–g4 are genetic influences on tree growth processes and Et is the error term incorporating all of the signal that is not controlled for by the other variables.

Given a non-stationary relationship between R and C today might simply mean that we have not taken enough into consideration all the other terms contributing to tree growth (under the assumption of a similar level of adaptation of the specific genotypes to their environments today and in the past). Let us for a second assume that we DO understand contemporary tree growth completely (i.e. we can quantify all terms in the equation), the correct application of the “uniformitarian principle” would be to assume that IF these conditions (the terms in the equation) had also happened in the past, the result (i.e. the tree ring) would have been the same as today. Nothing more – and nothing less. Non-stationary tree growth climate relationships are therefore definitely not a violation of the “uniformitarian principle”, but rather more likely related to our missing knowledge.

At this point, some words have to be spent on the possible “no-analogue problem” (Williams and Jackson, 2007). This problem, which is important for any paleo-science, states that environments change and combinations of environmental factors can vary in a way that there either existed a system in the past (or will in the future) for which we have no modern analogue (e.g. the Mammoth steppe of the last glacial maximum), or that we might encounter a modern system today, which is so different from the past that we cannot use it to draw inductive inferences about the past (Knight and Harrison, 2014). This is important to note as several authors have used, or at least implied, the “no analogue” issue as a possible explanation for a deviation from the “uniformitarian principle” (e.g. Fritts and Swetnam, 1989, Briffa et al., 1998). While it is true that the “no-analogue problem” does influence model validation statistics, and that inferences about the past from trees growing under atmospheric concentrations of >400 ppm in today’s nutrient- and pesticide-rich landscapes are not as straightforward as originally imagined, it is also true that there is NEVER a complete analogue and that we have, as usual in the natural sciences, to deal with uncertainties surrounding our estimates. A “no-analogue” problem does not violate the “uniformitarian principle”, it makes estimates more noisy.

For our tree-ring community we see several options to avoid further confusion. First and foremost the community should acknowledge that: The “uniformitarian principle” is “An undemonstrable, though entirely necessary, procedural assumption asserting spatial and temporal invariance of law’s describing nature’s processes” (Gould, 1965a). It has nothing to do with testable hypotheses, but is rather an assumption we agree on before we start with any scientific analysis. Once we accept this, there are two options (choose one):

  • Follow Gould one step further and assume that “the concept embodied in the procedural assumption is essential, but the term ‘uniformitarianism’ (in our case “uniformitarian principle”, italic inset by the author) is unnecessary, since the invariance assertion is synonymous with the principle of induction” (Gould, 1965a). Since all of natural sciences are based on inductive reasoning (and/or its weaker form of abductive reasoning), we do not need to name and list the principle any longer (Shea, 1982).

  • Retain the “uniformitarian principle” as a reiterating backbone of our scientific approach, a specific application of the general inductive concept to paleo-sciences in general and dendrochronology in particular.

The authors personally lean towards option one as the logical consequence. However, we do see a value in retaining the principle for teaching purposes. If we as a community chose option two, we can simply define:

The “uniformitarian principle” is an a priori assumption of spatial and temporal invariance of law’s describing nature’s processes. Applied to the dendro-sciences it means that the principle of aggregate tree growth is valid in time and space.

Section snippets

Acknowledgements

We thank two anonymous reviewers for their helpful input. This study is a contribution to the Virtual Institute ICLEA (Integrated Climate and Landscape Evolution Analysis) funded by the Helmholtz Association and the research consortium BaltRap (The Baltic Sea and its Southern Lowlands: Proxy-Environment interactions in times of rapid changes) funded by the Leibniz Association.

References (36)

  • K.R. Briffa et al.

    Trees tell of past climates: but are they speaking less clearly today?

    Philos. Trans. R. Soc. B

    (1998)
  • G. Camardi

    Charles Lyell and the uniformity principle

    Biol. Philos.

    (1999)
  • W. Cardwell

    Basic dendrochronology

    (2004)
  • M. Carrer et al.

    Long-term change in the sensitivity of tree-ring growth to climate forcing in Larix decidua

    New Phytol.

    (2006)
  • E.R. Cook

    The decomposition of tree-ring series for environmental studies

    Tree-Ring Bull.

    (1987)
  • I. Dorado-Liñán et al.

    Potential to explain climate from tree rings in the south of the Iberian Peninsula

    Clim. Res.

    (2012)
  • H.C. Fritts et al.

    Dendroecology: a tool for evaluating variations in past and present forest environments

  • H.C. Fritts

    Tree Rings and Climate

    (1976)
  • Cited by (15)

    • Stationarity of climate-growth response is only marginally influenced by the soil moisture regime in Western Siberia

      2021, Dendrochronologia
      Citation Excerpt :

      We had originally theorized that if drought stress might influence the temperature-growth response, as has been hypothesized for boreal forests (Barber et al., 2000; Beck et al., 2011; Seo et al., 2011; Franke et al., 2017), it might also influence stationarity (e.g., see D’Arrigo et al., 2008; Juday et al., 2015; Lange et al., 2020). Therefore, non-stationarity should be more common on dry micro-sites in these northern boreal regions, where a shift from temperature to moisture control is more likely (Wilmking et al., 2017; Peltier and Ogle, 2020). Our study, however, could not substantiate this hypothesis – micro-sites did not seem to affect stationarity in these northern boreal environments in any systematic way.

    • Non-linear regional weather-growth relationships indicate limited adaptability of the eastern Baltic Scots pine

      2021, Forest Ecology and Management
      Citation Excerpt :

      The secondary role of summer moisture (precipitation) regime (Fig. 3F, G, I) might be related to climatic changes and intensifying water deficit in temperate and hemiboreal forests (Harvey et al., 2020; Lloyd et al., 2013; Ohse et al., 2012; Friedrichs et al., 2009). This also implies interactions between the effects of winter temperature regime and summer water deficit (Fei et al., 2017; Matisons et al., 2017; Ohse et al., 2012), which diversify growth responses (Hofgaard et al., 2019; Wilmking et al., 2017). As hypothesized, the prevailingly non-linear weather-TRW relationships were revealed by the generalized additive model (Fig. 3), indicating presence of regional weather drivers of radial increment (Cavin and Jump, 2017; Restaino et al., 2016; Matias et al., 2017; Lloyd et al., 2011, 2013), which however, were similar to those identified by the linear correlation analysis (Fig. 2).

    View all citing articles on Scopus
    View full text