“Soft” forensic engineering investigation and peer review

I recently saw reference online to soft science and hard science. I was researching topics while also looking at titles of blogs I had posted in the past. My research was in the nature of an assessment of what more needs to be known about forensic engineering investigation and expert services.

I was troubled by the adjective “soft” in the reference and it not reflecting well on forensic work. Particularly a forensic investigation that had not been peer reviewed.

I found a lot of comment in the literature on what’s soft science as distinct from hard science. In the engineering spirit of trying to keep it simple I settled on the following:

  • Hard science relies on math and has more control of the variables and conclusions in applying the scientific method
  • Soft science relies on observational/empirical data and has less control of the variables and conclusions

If science can be soft and engineering is applied science can it be said there is soft and hard engineering? Would engineering relying on the Observational Method be an example of soft engineering? (Refs 1, 2 and 3)

Why do I think this is relevant? Because, to the uninformed soft engineering might not be seen as engineering. Does that mean soft science is not science? I don’t think so; check out the discussion at Wikipedia.

But, if nothing else, perceiving engineering as soft opens it up to hostile rebuttal review. That’s all the more reason for peer review particularly when a forensic engineering investigation is of a soft engineering failure.


Examples of soft engineering:

  • Investigation of the cause of a personal injury accident like slip and fall accidents
  • Forensic engineering investigation of failures, particularly of structures in the natural environment – in and on the ground – compared to those in the built environment – above the ground
  • Design of shoring for an excavation
  • Design and construction of foundations, particularly on soft and loose soils
  • Flood water protection
  • Design and construction of stable cut slopes. (A cut slope is a slope in natural, undisturbed soil, like along the side of a highway)


  1. One forensic observation does not a cause make Posted July 18, 2023
  2. Observational Method: Example #1 Posted July 31, 2023
  3. Observational Method: Example #2 Posted August 29, 2023

(Posted by Eric E. Jorden, M.Sc., P.Eng. Consulting Professional Engineer and Forensic Engineer, Geotechnology Ltd., Halifax, Nova Scotia, Canada, May 28, 2024 ejorden@eastlink.ca)    


Huffing and puffing about the number of methods for analysing data

Excuse me if I huff and puff a bit about being right. There really is only one way to analyse evidence in a crime to identify the bad guy – not four ways. Similarly, there is only one way to analyse data to learn the cause of an accident or failure in the built or natural environment.

This came out in a talk with a friend about his career in police work. He saw four ways of analysing data and evidence. I allowed maybe four but only one underlying method as posted in a previous blog. (Fig. 1)

My friend sees the four methods as follows:

  1. Tactical (Specific to a project or occurrence)
  2. Strategic (Addressing general issues, trending or potential) i.e. foresight, threat and risk assessments
  3. Intelligence (Discover plans or developments of concern)
  4. Business (As above, to monitor intelligence pertaining to the industry, company,  competitors, risks of takeover)

I see four different types of data and evidence in my friend’s list. I don’t see four different types of analytical method. Analysis is analysis is analysis – check out the definition in the dictionaries:

Four dictionaries – Cambridge, Brittanica, Merriam-Webster and Oxford -gave similar definitions of analysis. I’ve summarized these in the following as applies to forensic engineering investigation, and for sure as applies to catching the bad guys:

Analysis is the act of studying or examining “something” closely and carefully in detail, it’s elements or structure, to discover or understand more about its nature and the relationship of it’s parts, it’s essential features

The “something” could be a single task in a forensic engineering investigation and if it indicates the cause of an accident or failure in the built or natural environment, or supports the findings of another task.

(A little aside, there is a strong moral and ethical burden on an analyst to be thorough, accurate, unbiased, and objective. The result of the analysis must be credible and reliable and as much as possible, beyond reproach)

There was nothing in the four dictionaries about four methods of analysis. The definitions were very similar as I summarized above. Go see in the dictionaries I listed above – two of these were checked by my friend and two by me.

My friend may have been led astray by the nature of his work. He went on to list six (6) different purposes for an analysis, four (4) methods of analysis – see his list and my comments above – and 11 methodologies of analysis.

I’m not surprised at his detailed lists – like said in my earlier blog he did good work and travelled extensively. The lists help my friend and his colleagues do good police work – even if the four (4) types of data are mislabeled as four (4) methods.

The lists reminded me that I hadn’t seen comparable lists in forensic engineering investigation. If they haven’t been prepared they could be – the information is out there and followed in engineering work. Maybe the lack of published lists in engineering takes the huff and puff out of my sails.


  1. How many different ways can data and evidence be analysed? Posted May 17, 2024

(Posted by Eric E. Jorden, M.Sc., P.Eng. Consulting Professional Engineer and Forensic Engineer, Geotechnology Ltd., Halifax, Nova Scotia, Canada, May 16, 2024 ejorden@eastlink.ca)