Short communication: (What) To teach or not to teach – that is the question

There has been an explosion of knowledge in geotechnical engineering, yet the art and science paradigm remains deeply rooted in the field. Geotechnical engineering students will face today’s prevailing difficulties as well as new and demanding geotechnical challenges such as those associated with energy and the environment. Confronting these situations and needs requires a sound and resourceful foundation. Therefore, geotechnical engineering teachers must review their educational programmes (a) to reconsider the role of empiricism, (b) to prune incorrect concepts and biases (e.g., enduring misnomers, incorrect explanations, superseded graphical approaches, unsound tricks and fragile correlations, and education based on extremes (dry–saturated, clay–sand, drained–undrained)), and (c) to promote a careful understanding of fundamentals (e.g., the particulate nature of soils and fractured rocks, formation history, the essential relevance of effective stress, thermo-hydro-chemo-mechanical coupled processes, repetitive loads and ubiquitous localisations). Finally, they must continue reflecting on the role of the engineer in society, within an ever-changing world as the driver for innovation.


Introduction
The first century of modern geotechnology is coming to an end. There has been an explosion of knowledge in the field, and a great insight into soil and rock behaviour through exceptional experimental and numerical capabilities has been gained. Today, the field is broader than any single person can master, and geotechnical engineers publish in more journals than anyone can follow (probably more than 30 journals, complemented by a similar number of trade magazines). There is unlimited access to information, and effective measurement systems and powerful analysis/design tools are readily available.
The first century has also highlighted difficulties in geotechnical design, to the point that geotechnical engineers have extensively accepted that the field is a combination of both art and science. Could this be the reason for consistently disappointing blind prediction exercises? There is a need to reassess the role of empiricism, question often invisible biases and incorrect concepts or approaches that have crept into teaching and practice, and bring back an emphasis on deep understanding coupled with multidisciplinary fluency.
Research and education are warranted by difficulties and challenges in geotechnical engineering practice. Current students of the field will reach the prime of their professional careers at the time when geotechnical engineering challenges that are sprouting today will dominate their professional lives, for instance, issues related to climate change, sea level rise, energy needs and sustainability. Confronting those needs will require a sound and resourceful scientific foundation. In this context, geotechnical engineering teachers must pause to reflect on their educational programmes.
Bare core: the fundamentals While knowledge has expanded, teaching time remains limited and only a very small subset of concepts can be covered in the curriculum. The author often wonders about the key fundamental concepts that students should deeply understand if academic programmes were forced to reduce the contents of the courses to a bare minimumsay, a few pageswhile satisfying multidisciplinary fluency. The underlying assumption is that young engineers will be able readily to add pragmatic procedures to a well-funded knowledge structure while still preserving the engineering versatility fundamentals provide.
For the sake of this briefing, the author will address the last of these. The course on soils starts with the fundamental realisation that soils are particulate materials. Then, all observations summarised in Table 1 are true and causally related. The particulate nature of soils is recognised in the early stages of soil mechanics, yet it is seldom emphasised in class. Observations in the table apply to all soils and fractured rocks.

Not to teach (pruning)
There are enduring misnomers, superseded practices and restraining recipes in the field. Those in the geotechnical engineering community all share responsibility for pruning these out; in particular, educators, journal editors and conference organisers can play an effective role to this end. The following are examples.
■ Terms with multiple semantics. The term 'clay' is used to indicate size, mineral (crystal or amorphous), or any soil that plots above the A-line on the plasticity chart. Use the term 'clay' to indicate clay minerals, and classify fine-grained soils according to their sensitivity to changes in pore fluid chemistry. ■ Misnomers. The salient example is 'cohesive soil'this is a dangerous oxymoron indeed. Abandon the qualifier Soils are particulate geomaterials, therefore: Soils are inherently non-linear (Hertz and electrical contacts), non-elastic (Mindlin contact), porous and pervious (i.e., porosity between grains is interconnected).
Particle-level characteristics and processes integrate to make the macroscale response ■ size: determines the balance between particle-level forces (capillary and electrical forces gain relevance when particles are smaller than 10-50 μm) ■ shape: reflects formation history and affects grain packing, anisotropy, stiffness, strength and permeabilityamong others ■ spatial arrangement of grains: determined by electrical forces in fine-grained sediments (pore fluid pH and ionic concentration) and by grain shape and relative grain size in coarse-grained soils ■ porosity: varies widely and is stress dependent in (unstructured) fine-grained soils, but varies in a narrow range and is mostly defined at the time of packing in coarse-grained soils.
■ Both are very different: the key is to anticipate their distinct responses to imposed boundary changes. ■ The skeleton and the fluid interact: this gives rise to coupled fluid pressure, effective stress, volumetric strains and shear response. ■ Mixed fluids add capillary effects onto the particulate skeleton. ■ Generalisation: all hydro-chemo-bio-thermo-mechanical processes are coupled.
The mechanical behaviour of the particulate skeleton is effective stress dependent. ■ This includes stiffness (Hertz), frictional strength (Coulomb), and dilation/contraction upon shear (Taylor). ■ Frictional strength limits the maximum stress anisotropy a soil can experience. ■ Other properties may depend on effective stress as well (e.g., thermal conductivity of dry soils).
Particle-level deformation mechanisms change with strain level. ■ Small-strain deformation. It takes place at constant fabric and grain deformations concentrate at interparticle contacts; in this strain regime, volume change, pore pressure generation and frictional losses are minimal. ■ Large-strain deformation. It involves fabric changes (the role of contact-level grain deformation vanishes). ■ Threshold strain between the two regimes. It is higher for smaller particles and at higher confinement.
Soils are not inert and often change within the time scale of engineering projects.
■ Corollary: natural soils may behave differently from freshly remoulded clay or recently packed sand. 'cohesion', and related expressions such as 'cohesive soil' and 'cohesionless soil'. ■ Incorrect concepts. The term 'lubrication' is often misused in discussions of friction and explanations of the dry branch of the compaction curve. Primary and secondary consolidations are imagined as sequential rather than concurrent processes. Peak strength and critical-state void ratio are inferred even when specimens have experienced progressive failure and shear localisation. Tensile strength is invoked to explain desiccation crack formation. And the term 'thixotropy' is indiscriminately used in relation to time-dependent changes in fine-grained sediments. ■ Superseded. Still taught are graphical approaches that have become detached from their physical-mathematical underpinning (e.g., the determination of coefficient of consolidation using sqrt-t or log-t methods: a simple spreadsheet calculation can readily fit the diffusion equation to the data and incorporate other effects such as the change of permeability with void ratio). Teachers keep concepts developed for hand plotting and hand calculations (e.g., meaning of preconsolidation pressure and its determination), and preserve the use of parameters that add limited information (e.g., plastic limit (PL) is highly correlated to liquid limit for clay-dominant soils; then, we need to reassess its value and the adequacy of plasticity index-based correlations). ■ Restrictive/simplistic tricks that are not sound. From 'buoyant unit weight' and unconsolidated-undrained tests to total stress analysesparaphrasing J. Atkinson, are we not ready for a clean parting from total stress yet? ■ Fragile correlations and equations with local validity.
Diagnostic symptoms include dimensionally inhomogeneous expressions and equations that violate asymptotic trends for extreme values of the variable (e.g., the linear e-log s¢ equation). Instead, focus on physics-inspired correlations to attain more robust equations that satisfy asymptotic conditions.

To tweak and refocus
In an attempt to bring clarity, soil conditions have been polarised and a curriculum around extremes has been developed. Consider two cases. First, soils are taught as if they are either dry or water saturated, while reality involves these two extremes and all unsaturated conditions in between. Second, drained and undrained analyses are covered, yet these are two extremes when the rate of pore pressure dissipation is either much faster (drained) or much slower (undrained) than the rate of loading.
There are some old-sounding but most elegant concepts that the author likes to cover in class, but with renewed emphasis on understanding rather than on devoting full focus to the development of engineering solutions. Examples include 'feeling' equilibrium with Mohr circle, combining equilibrium and failure conditions in a Mohr-Coulomb analysis to show limiting stress anisotropy in soils, the elegance of elastic solutions, the 'essential engineering nature' of upper and lower bounds (i.e., what is needed is not the exact solution, but reliable and narrow bounds), and even flownets (i.e., to highlight the identification of boundary conditions and to 'experience' flow patterns and seepage forces).

To teach (evolving emphasis)
Gradually, the author's lectures are evolving by shifting emphasis and incorporating other topics thatin his mindstudents will need in order to keep up to date and intellectually agile in a changing world.
■ Place continued emphasis on the particulate nature of soils and fractured rocks, and the critical relevance of effective stress! Extend the coverage of other needed scientific foundations.
(Note: early reviewers of this briefing suggested a dual track: one that emphasises skills that will be immediately useful in practicewith exposure to the scientific foundationsand another that emphasises the scientific foundations and provides less of the practice-oriented skills.) ■ Present an updated discussion of formation history and diagenesis (with proper coverage of residual soils), bonding and structured soils (i.e., salient deviations from observations made in Table 1), natural and manufactured soils (e.g., mine tailings and fly ash), stratigraphy and spatial variability (at all scales). ■ Increase emphasis on well-designed field tests to measure properties for engineering design. Principally, the laboratorymeasured stiffness/compressibility is critically affected by sampling effects and aggravated by seating effects; in particular, oedometer tests should be limited to situations where the anticipated vertical shortening is a significant fraction of the specimen initial height. ■ Extend teaching examples to a wide range of fluids, pressure, effective stress and temperature conditions that upcoming geotechnical problems will impose (including grain-crushing effects). ■ Emphasise both short-and long-term performance monitoring with a focus on making predictions and assessing interpretations (noting that there are innate monitoring limitations such as accelerating bifurcations). ■ Increase awareness of the pervasive tendency to localisations of all kinds which break down the common assumption of homogeneity (from shear bands and compaction bands to dissolution pipes, flow localisation, fingering and a wide range of opening mode discontinuities or fractures). ■ Introduce repetitive loads (mechanical, thermal, chemical, moisture): they may determine long-term performance. ■ Continue reflecting on the role of the engineer in society, within an ever-changing world as the driver for innovation.

Closing
Geotechnical engineering has evolved and continues to develop as a result of the synergistic interaction between education, research and practice. This synergism is needed today more than ever before as the field gains pre-eminent roles in the most challenging problems humanity has ever faced. Students should be encouraged to embrace these challenges so they thrive in the new opportunities that the field presents, and let this excitement permeate into the classroom.
Geotechnical education has underplayed the importance of first principles while trying to focus on prescriptive solutions needed by practitioners to solve today's problems. If the pendulum is allowed to swing further toward the 'practice' end, advances needed to enhance the solution of today's problems will be delayed and future generations of students will be ill prepared to tackle the challenges that the profession will face.
Geotechnical engineering teachers are challenged to make difficult choices about their curriculum. Furthermore, they are constrained by the evolving civil engineering programmes that contain them. They should lead the change and define their role.